git@vger.kernel.org mailing list mirror (one of many)
 help / color / mirror / code / Atom feed
From: Jeff King <peff@peff.net>
To: Yoichi Nakayama <yoichi.nakayama@gmail.com>
Cc: Yoichi Nakayama via GitGitGadget <gitgitgadget@gmail.com>,
	git@vger.kernel.org
Subject: Re: [PATCH v4 2/2] git-jump: invoke emacs/emacsclient
Date: Thu, 24 Nov 2022 11:29:39 -0500	[thread overview]
Message-ID: <Y3+b80Vb2/6lKQ40@coredump.intra.peff.net> (raw)
In-Reply-To: <CAF5D8-u14grTa9cUo=ge8TVRKNYtVHWhFHDW8yYD9gD8=f0Byg@mail.gmail.com>

On Thu, Nov 24, 2022 at 09:32:45PM +0900, Yoichi Nakayama wrote:

> > Good point. The vim version is easier to fix (we just need to
> > double-quote \$1 inside the eval), but the fact that nobody has
> > complained is an indication that it does not really matter.
> 
> I've confirmed the vim version is fixed by
>     eval "$editor -q \"\$1\""
> 
> With your hint, I found the emacs version can be fixed
> by single-quoting the variable (I found a mistake in the
> emacs version. Since there is only one argument, I
> should use $1 instead of $@. I'll fix it.), and the vim
> version can be also in the similar form with single quote:
>     eval "$editor -q '$1'"

This isn't quite a full solution, though. The "$1" is expanded by the
outer double-quoted string, which is then fed to eval. The inner
single-quotes make most characters literally except for single-quotes
themselves. So if $1 has single-quotes, the eval will barf due to bad
syntax.

> The original vim version used the notation \$1 instead of $1.
> I'm worried that the emacs version might need the backslash.
> What does the backslash mean? Is it necessary?

As you figured out in the other email, it inhibits the outer layer of
expansion, and lets the eval expand it. This is the easiest way to pass
things through levels of shell evals (since otherwise you have to
actually quote, which is a real pain).

None of this is sufficient for your emacs example, though. There you
have three levels of quoting:

  - getting the argument intact into the eval; this can use the "\$"
    trick

  - the argument then appears inside a double-quoted string which will
    be evaluated by emacs. You'd want to protect it against
    double-quotes and presumably backslashes.

  - emacs will then execute the final command, presumably you a shell.
    So you'd want to protect against expansion in that shell. The
    easiest way to do that is usually to wrap each argument in
    single-quotes, and quoting against interior single quotes (by ending
    single-quote, adding a single backslashed single-quote, and then
    re-opening the single-quote). It's horribly ugly, but is (AFAIK) the
    shortest way to quote shell arguments, and what we usually do in
    Git.

Those are the three tricks I sent in the earlier email (though looking
at it again, I think the single-quote bits need to come first, so their
backslashes are then quoted to protect against emacs evaluation).

It's all quite confusing, which is why I am OK with just skipping it for
now. ;) The nice thing, though, is that doing the quoting right means
it's safe to get rid of the "cat", which solves your race problems in a
more direct and robust way.

-Peff

  parent reply	other threads:[~2022-11-24 16:29 UTC|newest]

Thread overview: 68+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-11-19 14:02 [PATCH 0/2] git-jump: support Emacs Yoichi NAKAYAMA via GitGitGadget
2022-11-19 14:02 ` [PATCH 1/2] git-jump: add an optional argument 'stdout' Yoichi Nakayama via GitGitGadget
2022-11-19 14:02 ` [PATCH 2/2] git-jump: invoke emacsclient Yoichi Nakayama via GitGitGadget
2022-11-19 15:53   ` Eric Sunshine
2022-11-19 23:44     ` Yoichi Nakayama
2022-11-19 23:59       ` Eric Sunshine
2022-11-21  4:05         ` Yoichi Nakayama
2022-11-20  1:27 ` [PATCH v2 0/2] git-jump: support Emacs Yoichi NAKAYAMA via GitGitGadget
2022-11-20  1:27   ` [PATCH v2 1/2] git-jump: add an optional argument 'stdout' Yoichi Nakayama via GitGitGadget
2022-11-21  5:43     ` Junio C Hamano
2022-11-21 11:25       ` Yoichi Nakayama
2022-11-21 18:18       ` Jeff King
2022-11-20  1:27   ` [PATCH v2 2/2] git-jump: invoke emacsclient Yoichi Nakayama via GitGitGadget
2022-11-21 12:26   ` [PATCH v3 0/2] git-jump: support Emacs Yoichi NAKAYAMA via GitGitGadget
2022-11-21 12:26     ` [PATCH v3 1/2] git-jump: add an optional argument '--stdout' Yoichi Nakayama via GitGitGadget
2022-11-21 18:38       ` Jeff King
2022-11-21 23:38         ` Junio C Hamano
2022-11-22 13:00           ` Yoichi Nakayama
2022-11-22 18:23             ` Jeff King
2022-11-22 13:29         ` Yoichi Nakayama
2022-11-21 12:27     ` [PATCH v3 2/2] git-jump: invoke emacs/emacsclient Yoichi Nakayama via GitGitGadget
2022-11-21 18:50       ` Jeff King
2022-11-22 12:06         ` Yoichi Nakayama
2022-11-22 14:18     ` [PATCH v4 0/2] git-jump: support Emacs Yoichi NAKAYAMA via GitGitGadget
2022-11-22 14:18       ` [PATCH v4 1/2] git-jump: add an optional argument '--stdout' Yoichi Nakayama via GitGitGadget
2022-11-22 18:30         ` Jeff King
2022-11-22 14:18       ` [PATCH v4 2/2] git-jump: invoke emacs/emacsclient Yoichi Nakayama via GitGitGadget
2022-11-22 16:40         ` Phillip Wood
2022-11-23  5:01           ` Yoichi Nakayama
2022-11-22 18:54         ` Jeff King
2022-11-23  5:33           ` Yoichi Nakayama
2022-11-24  1:09             ` Jeff King
2022-11-24 12:32               ` Yoichi Nakayama
2022-11-24 12:58                 ` Yoichi Nakayama
2022-11-24 16:31                   ` Jeff King
2022-11-24 16:29                 ` Jeff King [this message]
2022-11-25  3:46                   ` Yoichi Nakayama
2022-11-23  7:04       ` [PATCH v5 0/3] git-jump: support Emacs Yoichi NAKAYAMA via GitGitGadget
2022-11-23  7:04         ` [PATCH v5 1/3] git-jump: add an optional argument '--stdout' Yoichi Nakayama via GitGitGadget
2022-11-23  7:04         ` [PATCH v5 2/3] git-jump: move valid-mode check earlier Jeff King via GitGitGadget
2022-11-23  7:04         ` [PATCH v5 3/3] git-jump: invoke emacs/emacsclient Yoichi Nakayama via GitGitGadget
2022-11-23 14:58           ` Phillip Wood
2022-11-23 21:54             ` Yoichi Nakayama
2022-11-24  1:11         ` [PATCH v5 0/3] git-jump: support Emacs Jeff King
2022-11-24  3:47         ` [PATCH v6 " Yoichi NAKAYAMA via GitGitGadget
2022-11-24  3:47           ` [PATCH v6 1/3] git-jump: add an optional argument '--stdout' Yoichi Nakayama via GitGitGadget
2022-11-24  3:47           ` [PATCH v6 2/3] git-jump: move valid-mode check earlier Jeff King via GitGitGadget
2022-11-24  3:47           ` [PATCH v6 3/3] git-jump: invoke emacs/emacsclient Yoichi Nakayama via GitGitGadget
2022-11-25  3:36           ` [PATCH v7 0/3] git-jump: support Emacs Yoichi NAKAYAMA via GitGitGadget
2022-11-25  3:36             ` [PATCH v7 1/3] git-jump: add an optional argument '--stdout' Yoichi Nakayama via GitGitGadget
2022-11-25  9:06               ` Ævar Arnfjörð Bjarmason
2022-11-27  0:31                 ` Yoichi Nakayama
2022-11-25  3:37             ` [PATCH v7 2/3] git-jump: move valid-mode check earlier Jeff King via GitGitGadget
2022-11-25  3:37             ` [PATCH v7 3/3] git-jump: invoke emacs/emacsclient Yoichi Nakayama via GitGitGadget
2022-11-25  8:55               ` Ævar Arnfjörð Bjarmason
2022-11-25 16:01                 ` Yoichi Nakayama
2022-11-25 23:52                   ` Ævar Arnfjörð Bjarmason
2022-11-28  5:10                     ` Jeff King
2022-11-28 10:54                       ` Ævar Arnfjörð Bjarmason
2022-11-28 15:38                         ` Yoichi Nakayama
2022-11-27  0:37                   ` Yoichi Nakayama
2022-11-27  1:18             ` [PATCH v8 0/3] git-jump: support Emacs Yoichi NAKAYAMA via GitGitGadget
2022-11-27  1:18               ` [PATCH v8 1/3] git-jump: add an optional argument '--stdout' Yoichi Nakayama via GitGitGadget
2022-11-27  1:18               ` [PATCH v8 2/3] git-jump: move valid-mode check earlier Jeff King via GitGitGadget
2022-11-27  1:18               ` [PATCH v8 3/3] git-jump: invoke emacs/emacsclient Yoichi Nakayama via GitGitGadget
2022-11-28 10:13               ` [PATCH v8 0/3] git-jump: support Emacs Phillip Wood
2022-11-28 23:35                 ` Junio C Hamano
2022-11-29 21:05                 ` Yoichi Nakayama

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: http://vger.kernel.org/majordomo-info.html

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=Y3+b80Vb2/6lKQ40@coredump.intra.peff.net \
    --to=peff@peff.net \
    --cc=git@vger.kernel.org \
    --cc=gitgitgadget@gmail.com \
    --cc=yoichi.nakayama@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this public inbox

	https://80x24.org/mirrors/git.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).