A common challenge for beginners is distinguishing packages, modes, and
hooks, and since Doom's source code is meant to second as another layer
of documentation, I prefer to be explicit with hook symbols in
`add-hook!` calls (at least, for internal/module use).
At 0.1, Corfu pops up too aggressively and too often, even for fast
typists. This is especially disruptive with slow backends that might
block Emacs. I think 0.18-0.24 is the goldilocks zone, so I'll start
with 0.18, and maybe increase it later.
This is the only internal usage of :bind, which I want to avoid, because
it pulls in bind-key, which offers no real benefit over just about any
other key binding function or macro, so I'd rather use what is already
available (like map!).
This tidy ups the corfu module to use the logic when a user enables
corfu +orderless without enabling vertico. Since that's probably a niche
situation, we only specify commit to pin to orderless in the vertico
module so that contributors aren't tripped up with trying to remember to
update orderless in two different code locations. Of course, if this
turns out to be a more popular use-case we can re-evaluate this logic.
The tramp string used to connect to the container uses the first
`container-runtime ps` argument as a username, but the first argument
returned is the container ID, not a username.
Close: #7674
Co-authored-by: bergmannf <bergmannf@users.noreply.github.com>
Co-authored-by: elken <elken@users.noreply.github.com>
Corfu doesn't support a nil corfu-separator in general, but
+corfu-smart-sep-toggle-escape, specifically, will throw a characterp
type error if the user hasn't enabled +orderless.
We previously implemented only consult/vertico as a target for export,
now we have all of them. It was necessary to use case-by-case
conditions, unfortunately, because other UIs have subtle quirks that
prevent a single generalized approach to work.
Ivy is almost compliant, but it needs beg and end to not be markers.
Helm doesn't replace `completion-in-region-function`, it expects to go
around the default `completion--in-region`. It's supposed to add the
advice by itself, but it's very unreliable, so we do the wrapping
manually.
Ido doesn't implement `completion-in-region` and its `completing-read`
is retricted to a list of strings as table, so we use default
`completion--in-region` with no bells or whistles.
This removes the old `&` separator for Vertico (does anyone use that
instead of just space?) in favor of escapable space and unifies
orderless config with Corfu. Also implements smart separator
insert/escape/reset on `C-SPC`
Co-authored-by: Liam Hupfer <liam@hpfr.net>
Add CAPFs from cape:
- `cape-dabbrev`;
- `cape-elisp-block`;
- `cape-file`;
Fix some CAPFs via cape:
- Make non-exclusive, purified and silent `pcomplete-completions-at-point`;
- Make non-exclusive and non-interruptable `lsp-completion-at-point`;
- Make non-exclusive `eglot-completion-at-point`;
- Make non-exclusive `comint-completion-at-point`;
Fix and improve keybindings:
- Smart `DEL`;
Add depth to CAPFs, allowing ordering to be adjustable.
Yasnippet is now properly integrated! A previosly-unset default has now
been given to `corfu-on-exact-match`. With snippets, it causes immediate
expansion upon single match by default, so we set it to nil and
recommend against changing it in the README.
This commit's primary goal is allowing use of
[Corfu](https://github.com/minad/corfu) as an alternative to
[Company](https://github.com/company-mode/company-mode). It introduces a
module under `:completion` for this purpose, plus some conditionals on
other relevant modules to toggle functionality like lsp back-ends and
[Cape](https://github.com/minad/cape) capfs for certain modes.
Other optional or miscellaneous features include:
- Corfu is enabled in the minibuffer if `completion-at-point` is bound;
- Support for displaying the completion's documentation on a secondary
popup;
- Support for terminal display if :os tty;
- Support for icons if +icons;
* feat(vertico): completion highlights a la ivy
Adds completion highlighting that works similarly to ivy/counsel's
one (which is enabled by default). It'll highlight enabled major/minor
modes and directories in a different face. On by default.
Ref: https://github.com/minad/vertico/wiki#candidate-display-transformations-custom-candidate-highlighting
* fix(vertico): major mode not being highlighted
The major mode was not being highlighted correctly;
it should work now that the buffer is set correctly in
`+vertico-highlight-enabled-mode`.
* fix(vertico): make font lock prioritize match over type
The mode and directory highlights were [overriding the match
font-lock](https://github.com/doomemacs/doomemacs/pull/7706#issuecomment-1977722188).
This should resolve that by prioritizing the match font lock using
`'append` on `add-face-text-property` instead of `propertize`.
We should always bind to both "TAB" and "<tab>" - if they have been
mapped to different commands anywhere, Emacs will no longer treat them
as equivalent.
BREAKING CHANGE: This deprecates the IS-(MAC|WINDOWS|LINUX|BSD) family
of global constants in favor of a native `featurep` check:
IS-MAC -> (featurep :system 'macos)
IS-WINDOWS -> (featurep :system 'windows)
IS-LINUX -> (featurep :system 'linux)
IS-BSD -> (featurep :system 'bsd)
The constants will stick around until the v3 release so folks can still
use it -- and there are still some modules that use it, but I'll phase
those uses out gradually.
Fix: #7479
Depending on the system, this can cause sorting to be disabled long
before the performance hit would be felt, and it's not obvious that this
is the reason.
Currently, fd does not expose a way to match against only the path
components beneath the target path. When --full-path is specified, the
pattern matches against all components. For example, executing
consult-fd from /home/hlissner/.emacs.d with `home` as the query would
match every file (not excluded by other arguments) under .emacs.d.
Despite this --full-path behavior, fd still outputs relative paths, so
the user cannot even determine why some candidates are returned.
Until there is a method to match only against subdirectories, use
--absolute-path to at least to show the user why all matches are
occurring.
Ref: https://github.com/sharkdp/fd/issues/839
We were using `shell-quote-argument`, which is meant for passing file
names, strings and so on, not command-line arguments. For example,
`(shell-quote-argument "--foo=bar")` yields "--foo\\=bar", which is
obviiously invalid unless we're trying to pass an option named '--foo\'.
At any rate, there is no quoting/escaping for shells in the default
value of `consult-ripgrep-args`, so it doesn't look like this is
something we need to do.
`consult-buffer` uses `recentf` to populate file candidates. It is not
uncommon to use `consult-buffer` as a single entry point to buffers,
bookmarks and recent files, effectively replacing `recentf` and
`consult-recent-file`.
To improve startup performance, Doom enables `recentf-mode` after the
first file is opened (0e851ace9b). When executing `consult-buffer` at
startup, `recentf-mode` won’t be enabled yet. Add it to the
`consult-recent-file` advice to ensure that can’t happen.
Unlike `consult-recent-file`, `consult-buffer` does have significant
functionality without `recentf-mode`, but for the tiny fraction of Doom
users that disable `recentf-mode`, this is easy enough to
`advice-remove`.
Fix: https://github.com/doomemacs/doomemacs/issues/7461
These modules tend to conflict if more than one of them are enabled at
once. More systematic compatibility tests are in the works, but for now
this will do.
