Shouldn't hyper::header::Authorization be in headers instead of items?

[InstinctBas]

I'm getting a compilation error on this code:

let client = Client::new(rocket()).expect("valid rocket instance");
let mut response = client
	.get("/someurl")
	.header(header::Host {
		hostname: hostname.to_owned(),
		port: None,
	})
	.header(header::Authorization(
		header::Bearer { token: token.to_owned() }
	))
	.dispatch();

Which is:

the trait `std::convert::From<rocket::http::hyper::header::Authorization<rocket::http::hyper::header::Bearer>>` is not implemented for `rocket::http::Header<'static>`

If I understand the code here:
https://github.com/SergioBenitez/Rocket/blob/master/core/http/src/hyper.rs#L49
correctly, shouldn't Authorization be in import_hyper_headers! instead of import_hyper_items! ?

[jebrosen]

Unfortunately that wouldn't work. import_hyper_headers! generates a call to to_string(), and Authorization is one of only a few headers that does not implement Display.

Rocket could implement it manually for Authorization or change the approach entirely, but any fix is unlikely to be in a release before 0.5, which currently loses all typed header support. #1067 is the tracking issue to decide what to do with them going forward.

[InstinctBas]

Bugger... it's ashame we don't have fall back where you can add a header by supplying a simple name/value pair. Or is there an option for that which I've missed?

I've worked around it for now by sending the data as a cookie instead of in the header. Just trying to automate testing my API with CI :)

[jebrosen]

Or is there an option for that which I've missed?

You can use Header::new() with a name and value. Something like Header::new("Authorization", format!("Bearer {}", token)) should work.

[InstinctBas]

Sweet! that'll do :)

[InstinctBas]

@jebrosen hmm, it's complaining rocket::http::hyper::header::Header is a private member. Is there a different path to it?

[jebrosen]

I think that's not just a different name, it's the wrong Header. rocket::http::Header is the one you're after.

[InstinctBas]

Hmmm wonder where I got the other path from, anyway that worked :)

[SergioBenitez]

This is gone in master. #1067 tracks bringing something similar back. Ideally we'd resolve these kinds of issues then.