Custom C-i bindings do not play well with citar-open and RET

[dschrempf]

Describe the bug
In citar-open, pressing RET results in an error <C-i> is undefined.

To Reproduce

1. Set

  (define-key input-decode-map (kbd "C-i") [C-i])

2. Bind <C-i>, for example using general:

  (general-mmap "<C-i>" #'better-jumper-jump-forward)

3. Run citar-open and press RET.

Expected behavior
The entry is opened.

Screenshots
If applicable, add screenshots to help explain your problem.

Emacs version:
29.1

Context

• Most likely related to RETURN has no effect in citar minibuffer #707.
• Setting (setq citar--multiple-setup (cons "<tab>" "RET") is a workaround (but in my opinion it should not be necessary).
• Could be a problem of my setup (I try to distinguish between C-i and <tab>

[bdarcus]

Just in case, OS and (if relevant) distro and desktop?

[dschrempf]

Thanks for having a look!

+ nix-shell -p nix-info --run "nix-info -m"
 - system: `"x86_64-linux"`
 - host os: `Linux 6.1.54, NixOS, 23.11 (Tapir), 23.11.20230919.5ba549e`
 - multi-user?: `yes`
 - sandbox: `yes`
 - version: `nix-env (Nix) 2.17.0`
 - nixpkgs: `/nix/store/cz973qsjldbw4x7fx0rwcvhr9k645vz2-source`

I use XMonad.

[bdarcus]

When you get a chance, @aikrahguzar, any input?

[aikrahguzar]

This is exactly #707 and I don't know of any completely clean solution for that since it is a workaround for a limitation of completion ui. Changing the default value of citar--multiple-setup to use <tab> is the only option I can think of but I don't know if there are any unintended consequences.

[bdarcus]

I don't know of any completely clean solution for that since it is a workaround for a limitation of completion ui. Changing the default value of citar--multiple-setup to use <tab> is the only option I can think of but I don't know if there are any unintended consequences.

@dschrempf - thoughts?

Doesn't seem like there's much we can do about this?

[dschrempf]

Using the workaround is an option for me, but I do not know if this situation is desired.

I do not know the internals of citar (Does it use keybinding macros?), but in opinion, rebinding a key should not lead to an error in a seemingly independent library. If there is no way to fix this, so be it 🤷. Citar is a great package, thank you!, and I am happy that it works the way it does 😄.

[bdarcus]

@aikrahguzar did mention changing the default value as a possibility.

[dschrempf]

This would mean that citar does not anymore work as expected when Emacs is used from the terminal, is that so?

Also: The completion UI that you mentioned. Is this an external library?

[bdarcus]

This would mean that citar does not anymore work as expected when Emacs is used from the terminal, is that so?

IDK; that would obviously be a problem if so.

Also: The completion UI that you mentioned. Is this an external library?

I think he meant completing-read.