Re: Bug: git-p4 edit_template() and P4EDITOR w/options

From: Junio C Hamano <gitster@pobox.com>
Date: May 4, 2015, 22:01 UTC
Message-ID: <xmqqh9rsovd4.fsf@gitster.dls.corp.google.com>
In-Reply-To: <5D2E2EAF-FFE7-437F-A716-E152E865E634@pixar.com>

Chris Lasell <chrisl@pixar.com> writes:

Show 12 quoted lines

> PS: 'mate' is the CLI invocation of the TextMate GUI editor for OS
> X. The -w is required in this instance, or the 'mate' command
> would return instantly when the document opens. The -w causes it
> to wait until the document window is closed.
>
> I have noticed that the help output for mate says:
> =======
> By default mate will wait for files to be closed if the command name
> has a "_wait" suffix (e.g. via a symbolic link)
> =======
>
> and I have instructed my user to do just that for now.

I think that is not merely "for now" but is the way the command and the environment variable are designed to be used. A quick websearch for [$EDITOR environment with parameter] found this one

  http://superuser.com/questions/521070/unix-environment-variables-with-arguments

which seems to be talking about a similar issue (unrelated to Git).

The relevant part of git-p4 is this:

        # invoke the editor
        if os.environ.has_key("P4EDITOR") and (os.environ.get("P4EDITOR") != ""):
            editor = os.environ.get("P4EDITOR")
        else:
            editor = read_pipe("git var GIT_EDITOR").strip()
        system([editor, template_file])

It grabs $EDITOR (or $GIT_EDITOR) and treats it as the path to the editor executable, without letting shell to split that into words at whitespace boundaries, so that you can say things like

	EDITOR="/User/me/My Programs/nano"

The way we spawn EDITOR in our core codepaths matches what git-p4 does, too:

	const char *args[] = { editor, real_path(path), NULL };
	struct child_process p = CHILD_PROCESS_INIT;
	int ret, sig;

	p.argv = args;
	p.env = env;
	p.use_shell = 1;
	if (start_command(&p) < 0)
		return error("unable to start editor '%s'", editor);
	...

So...