Closed hai-nc closed 3 years ago
Maybe that can be done, yes. But I would need someone to do heavy testing for some time. I admit I don't know perfectly how keymap parents work or what the potential impacts would be of putting a parent keymap where there is none. So try it out it your copy of sly-mrepl
and let me know.
By the way, I am a heavy user of reverse i-search, but I call it with C-r (which is not regexp-capable).
I would need someone to do heavy testing for some time
I guess "heavy testing" means something like daily using the repl for programming for a few months, which I am not qualified for - I use it may be once or twice every couple of days just to quicklisp-install or to call some lisp functions only.
I think I will just customize sly-mrepl-mode-map in my own dot emacs file instead. It's less risky than changing sly-mrepl.
Thanks for the C-r tip. It's very handy but sometimes my use case is to access inputs from previous sly sessions or avoid losing focus from the current view, so I need M-r as well.
I think that's a good idea. You can also try (set-keymap-parent sly-mrepl-mode-map comint-mode-map)
in your emacs and see what that brings.
Also C-r
also lets you access inputs from previous sessions.
It used to be the case that in the sly REPL one can isearch input history by pressing "M-r" (I think it mapped to 'comint-history-isearch-backward-regexp'), but when I used sly-20210114.912 (on Emacs 27.1debian), "M-r" no longer map to the comint's function.
At a quick look at "/sly-20210114.912/contrib/sly-mrepl.el", line 183, I saw "(set-keymap-parent sly-mrepl-mode-map nil)" which likely be the cause.
I wonder if 'sly-mrepl-mode-map' can inherit the 'comint-mode-map' (by not having line 183) so that some convenient keybindings like "M-r" can be used as before?