Fix to allow any remote name to be used

[meltinglava]

Fixes #411

TODO: Is it ok if we take the first one we find. Are there any examples
of having more than one file in refs/remotes/origin/?.

[meltinglava]

AppVeyor build fails on linking. Looks like this is the build not these changes, as I found the same in other PR-s.

[ordian]

Thanks for the PR!

We would also need to fix the code that checkouts the registry

cargo-edit/src/fetch.rs

Lines 126 to 159 in 034f6ef

	let refspec = "refs/heads/master:refs/remotes/origin/master";
	fetch_with_cli(&repo, registry.as_str(), refspec)?;
	Ok(())
	}
	// https://github.com/rust-lang/cargo/blob/57986eac7157261c33f0123bade7ccd20f15200f/src/cargo/sources/git/utils.rs#L758
	fn fetch_with_cli(repo: &git2::Repository, url: &str, refspec: &str) -> Result<()> {
	let cmd = subprocess::Exec::shell("git")
	.arg("fetch")
	.arg("--tags") // fetch all tags
	.arg("--force") // handle force pushes
	.arg("--update-head-ok") // see discussion in rust-lang/cargo#2078
	.arg(url)
	.arg(refspec)
	// If cargo is run by git (for example, the `exec` command in `git
	// rebase`), the GIT_DIR is set by git and will point to the wrong
	// location (this takes precedence over the cwd). Make sure this is
	// unset so git will look at cwd for the repo.
	.env_remove("GIT_DIR")
	// The reset of these may not be necessary, but I'm including them
	// just to be extra paranoid and avoid any issues.
	.env_remove("GIT_WORK_TREE")
	.env_remove("GIT_INDEX_FILE")
	.env_remove("GIT_OBJECT_DIRECTORY")
	.env_remove("GIT_ALTERNATE_OBJECT_DIRECTORIES")
	.cwd(repo.path());
	let _ = cmd.capture().map_err(|e| match e {
	subprocess::PopenError::IoError(io) => ErrorKind::Io(io),
	_ => unreachable!("expected only io error"),
	})?;
	Ok(())
	}

[ordian]

I think it would be safer to try only known names master and if it fails, then HEAD

[meltinglava]

I think this is mostly ok. I wanted to know if there are any reference on how cargo deals according to this. As I think most users will never toutch this location I think it should be fine. If we want to check for names i would rather collect them into a vec and then first check master then HEAD then first element. Not sure about that order either.

[ordian]

Thanks! (sorry for the delay)

[ordian]

@killercup do you know what happened to travis CI? Can't merge the PR, tests seem to pass locally.

[killercup]

No idea, the build doesn't appear in the travis dashboard for some reason. I'll force merge it.

[killercup]

Github Action's CI now tells me some checks fail: https://github.com/killercup/cargo-edit/actions/runs/197551357

I dont have time to look into it right now, but it'd be cool if we can get a fix fo this merged soon.

[ordian]

ah, that's clippy and fmt, windows was failing before due to #377