Closed arkhan closed 4 years ago
Got the same issue! I think it was caused by this comment. Overall, I feel that the prettification is very fragile and will break repeated characters into prettified chunks as well which looks glitchy and distracting. I think some form of contextual parsing might be useful?
I've not had time to test anything yet, but I'm thinking that maybe using word-boundaries (so we have to start and end prettyficiation at boundaries) might be a good starting point.
For what it's worth, I set the fira compose predicate to prettify-symbols-default-compose-p
and it seems to work a lot better overall, including both comments and spurious wedged symbols.
@tgolsson thanks, setting the configuration as follows:
(leaf pcache
:straight t
:init
(setq pcache-directory (let ((dir (concat conf:cache-dir "pcache/")))
(make-directory dir t)
dir)))
(leaf unicode-fonts
:straight t
:init (unicode-fonts-setup))
(global-font-lock-mode 1)
(set-face-attribute 'default nil
:font "Fira Emacs Retina"
:height 115
:slant 'normal
:weight 'normal
:width 'normal)
(leaf fira-code
:straight (fira-code
:type git
:host github
:repo "johnw42/fira-code-emacs")
:setq (fira-code-compose-predicate . 'prettify-symbols-default-compose-p)
:hook (prog-mode-hook . fira-code-mode))
(leaf prog-mode
:config
(add-hook 'prog-mode-hook
(lambda ()
(mapc (lambda (pair) (push pair prettify-symbols-alist))
'(("in" . ?\u2208)
("IN" . ?\u2208)
("not in" . ?\u2209)
("NOT IN" . ?\u2209)
("not" . ?\u00AC)
("NOT" . ?\u00AC)))))
(defun indicate-buffer-boundaries-left ()
(setq indicate-buffer-boundaries 'left))
(add-hook 'prog-mode-hook #'indicate-buffer-boundaries-left)
(add-hook 'prog-common-hook
(lambda ()
(font-lock-add-keywords nil
'(("\\<\\(FIX\\|FIXME\\|TODO\\|BUG\\|HACK\\):" 1 font-lock-warning-face t))))))
But you should also change the "@" symbol
Got the same issue! I think it was caused by this comment.
Yes -- that was me! (Sorry, I should have realized beforehand that it would trigger that sort of behavior).
Overall, I feel that the prettification is very fragile and will break repeated characters into prettified chunks as well which looks glitchy and distracting. I think some form of contextual parsing might be useful?
I agree! I hadn't had a problem since I made that commend (and the commit was made by @johnw42), but I see what happens.
Now I have tested setting fira-code-compose-predicate
to 'prettify-symbols-default-compose-p
and it seems to work fine for most of my use cases.
Oh - I just pulled and commit 71b35183584f483e1d5957fa547e26c496eb6000 seems to fix the issue.
Oops, I'm just now seeing all these messages! I think everything is working now so I'm gonna close the issue. Feel free to file another one if more problems turn up.
Overall, I feel that the prettification is very fragile and will break repeated characters into prettified chunks as well which looks glitchy and distracting. I think some form of contextual parsing might be useful?
Oh, it's absolutely very fragile, and I keep discovering subtle new ways for it to be broken. I suppose I'll make a proper release once I'm satisfied I don't need any more API changes.
Great job, congratulations.
In the last commit you get a strange behavior as you can see in the attached image.
In my configuration I have it like this:
Thanks