Displaying images in Emacs in Kitty

[gzagatti]

I was wondering if anyone had any luck displaying images inside of Emacs inside of kitty. Is that even possible? If so, how can we do it?

I wanted to display inline images in org files and ein notebooks. Thus, it would be very useful for my workflow.

The function display-images-p suggests that this is not possible in kitty. But I'm wondering if there's a way around it.

[page-down]

It looks like this requires the solution discussed in the following link.

#4021

[kovidgoyal]

Far better would be for someone to add support for the full kitty graphics protocol to emacs. https://sw.kovidgoyal.net/kitty/graphics-protocol/

[gzagatti]

Thanks for pointing it out! I'm glad it's possible. Unfortunately, it's well beyond my capability 😞

[gzagatti]

I sort of got what I wanted by writing a custom kitten that will create a new window and then cat the image that I want to see from Emacs. In Emacs, I only need to set a custom mailcap-user-mime-data such that it automatically uses my kitten when opening an image in an external viewer.

The kitten looks like below. I'm not sure if that's the best way to send a command to the other window. For instance, nothing happens if the pointer in the other window is in normal mode. How could I make sure that the command simply gets sent to the shell and executed?

Also, boss.launch does not return a window, so I have to search for it just after creating it. However, the source code seems to suggest that a window is returned.

# imgviewer.py
from typing import List
from pathlib import Path
from kitty.boss import Boss
from kittens.tui.handler import result_handler

def main(args: List[str]) -> str:
    pass

@result_handler(no_ui=True)
def handle_result(args: List[str], result: None, target_window_id: int, boss: Boss) -> None:
    args = args[1:]
    files = []
    w = boss.window_id_map.get(target_window_id)
    for arg in args:
        file = Path(arg).expanduser()
        if file.exists():
            files.append(file)
    if len(files) == 0:
        return
    try:
        imgviewer = next(boss.match_windows(f"title:imgviewer{w.id}"))
    except StopIteration:
        imgviewer = boss.launch("--window-title", f"imgviewer{w.id}",
            "--cwd", "current", "--keep-focus")
        imgviewer = next(boss.match_windows(f"title:imgviewer{w.id}"))
        imgviewer.write_to_child("export HISTFILE=\r")
    imgviewer.write_to_child(f"timg {' '.join(str(f) for f in files)}\r")

And in Emacs, I add the following to the configuration file:

(custom-set-variables                                                                                                     
	'(mailcap-user-mime-data '(("kitty @ kitten ./imgviewer.py %s" "image/png" nil))))

This is how it looks like:

[kovidgoyal]

You dont need to do any of that if you just want to open the image in a new window. You can just do (mailcap-user-mime-data '(("kitty kitty +kitten icat --hold %s"))'

[gzagatti]

Thanks. I wanted to persist the window such that it becomes sort of a dump for any image I want to cat when during my session. This is the reason why I ended up developing the proof-of-concept kitten. It was also a good chance to learn how to create one 😄

[kovidgoyal]

Ah cool. Note that this can be done purely by using the remote control api without needing a kitten, but of course a kitten works as well.

[retpolanne]

Chiming in with a huge thread about implementing kitty protocols on emacs

https://lists.nongnu.org/archive/html/emacs-devel/2022-09/msg00384.html

[vonpupp]

I am a new user of Kitty. Be able to see inline images of emacs within Kitty would be just amazing!

[lf-araujo]

Once the mimecap custom variable is added, how to open the link? Clicking a link in orgmode still opens the binary within emacs.