M-x aya-open-line opens two new lines

[ardpp]

Open the scratch buffer and enable linum-mode to better see what I mean.

Put

(open-line 1)
(aya-open-line)

in the buffer.

Evaluation of any of those will add one new line as expected. However M-x aya-open-line adds two lines. Happens to me even with emacs -Q

[abo-abo]

Can't reproduce. Can you add more details?

[ardpp]

There is at least some inconsistency that I can reproduce with the following steps

• start emacs with emacs -Q (and stay in the scratch buffer)
• initialize packages via M-x package-initialize
• load auto-yasnippet
• enable lispy-mode

Put the following into the scratch buffer | denotes the cursor (open-line 1)| pressing e (for special-lispy-eval) opens a new line an the cursor stays in place. The same happens if you do M-x open-line or mark region and call eval-region

If you do the same with (aya-open-line)| After e a new line is opened an the cursor stays in place as before. When you do M-x aya-open-line (or mark the region an evaluate it) then you get

(aya-open-line)
|

Which looks as if open-line is followed by newline (With linum-mode enabled you see that two lines are actually added in the latter case.)

emacs and the packages are all up to date (from git resp. melpa)

[abo-abo]

Still can't reproduce. It could be an issue of you having different abbrevs and snippets than me. Maybe it's possible to try without those.

Also, try to edebug aya-open-line and see what code triggers.

[ardpp]

I have tried on a different machine with an empty .emacs.d directory except for a very minimalistic init.el that installs and loads auto-yasnippet. Also there I can reproduce the inconsistency. I add a gif that maybe explains better what is going on than I do in words. I hope that the details are visible enough. Basically in the scratch buffer I enable linum-mode. Write fi and then keep pressing C-o for aya-open-line

First C-o expands the snippet, second opens a new line. Subsequent calls of aya-open-line open two new lines which open-line does not

[abo-abo]

Can you provide this minimalistic init? I see the problem you have, but still can't reproduce.

[ardpp]

I created a new directory ~/.testemacs.d/ and the file init.el in that directory with the following contents

(defvar emacs-d "~/.testemacs.d")
(let ((default-directory "/usr/share/emacs/site-lisp/"))
  (normal-top-level-add-subdirs-to-load-path))

(setq package-user-dir (expand-file-name "elpa" emacs-d))

(require 'package)

(add-to-list 'package-archives '("melpa" . "http://melpa.org/packages/") t)
(add-to-list 'package-archives '("org" . "http://orgmode.org/elpa/") t)
(package-initialize)
(package-refresh-contents)

(unless (package-installed-p 'use-package)
  (package-install 'use-package))

(require 'use-package)

(use-package yasnippet
  :ensure t
  :diminish yas-minor-mode)

(use-package auto-yasnippet
  :ensure t
  :commands aya-create aya-open-line)

(global-set-key "\C-o" 'aya-open-line)

then starting emacs via emacs -q -l .testemacs.d/init.el and doing the things described in above posts reproduces the things for me.

[abo-abo]

I don't know if it matters in this case, but there's a difference between emacs -q and emacs -Q. The first one might still load custom code, e.g. /usr/local/share/emacs/site-lisp/. I recommend to always use emacs -Q.

[ardpp]

There seems to be something like a memory effect. Looking at the code of the function I do not understand why that happens. I can reproduce it on every machine that I have access to (Archlinux and Gentoo with emacs 24.5 and 25.1).

When I put in the directory ~/.testemacs.d/elpa/ only the yasnippet and auto-yasnippet directories and put in the ~/.testemacs.d/init.el the following code

(load-file "~/.testemacs.d/elpa/yasnippet-20160801.1142/yasnippet.el")
(load-file "~/.testemacs.d/elpa/auto-yasnippet-20160524.618/auto-yasnippet.el")
(global-set-key "\C-o" 'aya-open-line)

Then I can reproduce what is shown in the above gif even with emacs -Q -l .testemacs.d/init.el

[abo-abo]

Thanks, I was finally able to reproduce. The issue was that I was using the stable version of yasnippet. I've just fixed that bug that comes due to the interaction with the MELPA version.

[ardpp]

Thank you!