Reduce the width of the listed key bindings by:
Moving the range and highlighted symbols counter to the right of the transient
state title.
Shorten some key description names.
Reduce three instances of "search" to just one.
And reorder the listed keys based on their search scope, small to large:
swoop, buffers, files, project.
Add these keywords to macro spacemacs|add-company-backends
:append-hooks controls wether we add the hooks to major mode hook variable
:call-hooks conrols wether we execute the created hooks function right away
Basically for multiple backends setup we want to pass :append-hooks to nil and
:call-hooks to t.
Reverting a buffer could put multiple entries of the same backend in the buffer
local `company-backends` variable.
This is due to the fact that we mutate `company-backends-xxxx-mode` variable
with:
(setq ,backends-var-name
(mapcar 'spacemacs//show-snippets-in-company
,backends-var-name)))
Solution is to add a new variable that won't be mutated when we add the support
for snippets.
When multiple modes where provided to macro spacemacs|add-company-backends along
with variables, only the first mode was taken into account as the `variables`
variable was nil after the first interation on `modes`.
This reverts commit 29c78ce841 and all other fixes
that have been made afterwards.
The motivation is that use-package is seen by many as a replacement for
`require`. Is use-package always defer the loading of packages then is breaks
this use case, this does not respect POLA so even if it was making Spacemacs
loading faster (up to 3s faster on some startup on my machine) we just cannot
use it, it would be irresponsible. Spacemacs should be easy to use, loading
performance will come with time but it is not a priority.
Fix#10177
counsel-projectile has updated custom actions mechanism[1] and
counsel-projectile-switch-project-actions is no more defined
Also,
- moved counsel-projectile into spacemacs-layouts from ivy layer as
perspective/layout based project switching depends on it instead of swiper
- removed swiper from spacemacs-layouts
[1] a4e9a34d7f
Also remove unneeded addtion of yasnippet-snippets directory as it is done
automatically in the package.
push is to avoid in :init blocks because it will add duplicated entries when
the user reload the configuration with SPC f e R.
Had to create dummy init functions at some places since the owner of a package
is the last layer that defines the init function of a package. And a package
can be installed only if it has an owner.
abo-abo/swiper@4f57b5a9 removed the counsel--git-dir variable. Code
that used that variable is now expected to use (ivy-state-directory
ivy-last) instead. Update the code to reflect this change.
Signed-off-by: Chirantan Ekbote <chirantan@chromium.org>
When a layer doesn't have a package, its name was added to the candidate list as
a symbol instead of a string, breaking actions such as
layer-action-open-packages which expect a string.
In perspective transient state. b and l keys in the docstring have no
corresponding functions declared whenever both helm and ivy layers are not
used.
Add two private variables to fix the issue:
- spacemacs--persp-display-buffers-func
- spacemacs--persp-display-perspectives-func
These variables are set to the correct functions by the helm and ivy layers via
a use-package hook.
Default is `ignore` function so b and l does nothing if both helm and ivy layers
are not used, TODO: we should find a better default function.
Shadowing is now control by layer property ':can-shadow' only.
can-shadow is a commutative relation, if layer1 can shadow layer2 then layer2
can shadow layer1.
the shadow operator is a binary operator accepting two layer names, it is not
commutative and the order of the operands is determined by the order of the
layers in the dotfile (like the ownership stealing mechanism).
If ':can-shadow' is set explicity to nil in the dotfile then the layer won't
shadow any layer.
For instance to install both ivy and helm layer:
(setq dotspacemacs-configuration-layers
'(
ivy
(helm :can-shadow nil)
)
note that due to the commutative relation the above example can also be
written (in this case, ':can-shadow' should be read ':can-be-shawdowed'):
(setq dotspacemacs-configuration-layers
'(
(ivy :can-shadow nil)
helm
)
Layers can now declare in their layers.el file that they shadow one or more
layers using the following functions:
- configuration-layer/shadow-layers
- configuration-layer/shadow-layer
Those function are commutative so:
(configuration-layer/shadow-layer 'layer1 'layer2)
is the same as
(configuration-layer/shadow-layer 'layer2 'layer1)
and means that
layer1 shadows layer2
and
layer2 shadows layer1
The typical use-case is helm and ivy layers. Helm shadows the ivy layer and
Ivy shadows the helm layer.
Shadowing is sensitive to the order of declaration of layers in the dotfile,
for instance:
(setq dotspacemacs-configuration-layers '(
helm
ivy
))
means that ivy shadows helm so helm layer is effectively ignored,
whereas
(setq dotspacemacs-configuration-layers '(
ivy
helm
))
means that helm shadows ivy so ivy layer is effectively ignored.
This mechanism can be turned off using the :can-shadow keyword:
(setq dotspacemacs-configuration-layers '(
ivy
(helm :can-shadow nil)
))
means that both ivy and helm layers will be installed (not recommended in this
case)
Note that the `:can-shadow` mechanism will be fully implemented in a next
commit.
dotspacemacs-directory-snippets-dir contains a path to a directory with yas
snippets in ~/.spacemacs.d (or arbitrary picked dotspacemacs direcotry).
However, while it's not always present (user has to create it manually)
it is always checked while yas is loaded which generates errors.
This diffs forces checking whether the assumed path exists before adding
it to the yas-snippet-dirs.
This replaces the older pattern
:toggle (configuration-layer/package-usedp ..)
This implementation ensures that :disabled-for honors dependent packages, i.e.
if package a depends on package b, which is owned by layer c, and layer c is
disabled for layer d, then neither package a nor b will be configured for layer
d. Previously, this was only true for package a, but not b.
This commit also fixes:
- configuration-layer/describe-package now shows which post-init and pre-init
functions are disabled, if any
- Does not recreate all layer objects unconditionally when calling
configuration-layer/discover-layers. Previously, this led to all layers being
recreated after e.g. `SPC h SPC`, without any of the dotfile information.
Since this information is now necessary for
configuration-layer/describe-package, it’s important that we don’t clear the
indexed layers when invoking this function.
By convention, code markup (`~`) is reserved for keybindings in Org-based
documentation in Spacemacs. Verbatim markup (`=`) is reserved for code and
other code-like things. So change several readmes to reflect this convention.
Use verbatim markup for things like (non-exhaustive list):
- Emacs Lisp functions, modes, buffers, etc.
- Environment variables
- Directory paths
- Code in general
This is a follow-up commit after #8155 has been merged. It applies the same
changes to code of the `ivy` layer. It is not part of #8155 because there was
no `ivy` layer when that PR was submitted.
Add two new functions: `spacemacs/transient-state-register-add-bindings`
and `spacemacs/transient-state-register-remove-bindings` to prevent layer
authors and end users from dealing with the underlying variables' subtleties.
Emacs doesn't handle long line lengths very well. If you accidentally rg a file
a file that has very long line lengths it can freeze emacs. This prevents rg
from returning any results with more than 150 characters. Instead, it will
indicate that there were results in that file but they were elided.
The keybinding has changed from `SPC h l` to `SPC r l`; see the changelog, entry
0.200.0 (2016/10/02). So correct the documentation to reflect that.
`SPC h l` is currently bound to `helm-spacemacs-help-layers` (on branch
`develop`).
It is an Emacs-wide convention to use `RET` when referring to the act of
pressing the return key. Spacemacs seems to follow that convention as well. So
fix some Org files that were deviating from this convention.
In particular, replace usages of `<return>` and `return` with `RET`.
Also update the description of an `M-RET` keybinding[1] based on the docstring
of `ido-select-text`.
[1]: Suggested-by: bmag <bmag@users.noreply.github.com>
* Don't list imenu if it isn't loaded
* Add recentf config to :post-init function
* Add :init and :config sections for counsel and ivy for accessibility
* Add projectile alt actions to invalidate cache
* Add recentf alt actions to refresh and delete items
* Load flx since it is listed
Add a README.org to the Helm layer to fix `helm-spacemacs-help-layers`
(`SPC h l`). There is little actual documentation -- it just links to
the official docs as well as the Helm wiki and the Helm guide.
Issue #8679
Otherwise, the Haskell layer throws an eager expansion error because
`layers/+lang/haskell/funcs.el`, which calls this macro, is loaded before
`layers/+completion/auto-completion/config.el`, which defines
`spacemacs-default-company-backends`.
rg tries to be smart about printing line numbers in its output (if running
inside a TTY or not). On Windows, sometimes (always?) it chooses not to print
line number. However, we always want line numbers, so we need to pass -n flag.
Also fix a minor grammar mistake.
Getting FAQ candidates requires org to be loaded which in turn requires to
load magit (via orgit) etc...
This commit moves the FAQ helm source to a new feature helm-spacemacs-faq and
removes the FAQ source from SPC h SPC so the FAQs are only available as a helm
source via SPC h f.
But we have still the FAQ org file entry in SPC h SPC so it is easy to open
the file from the main help key binding.
Enabling a company backend for a specific mode was a tedious tasks with code
scattered at different locations, one for local variable definitions, one for
company hook function definitions and another where the backends were pushed to
the local variables (which was problematic, since we ended up pushing the same
backends over and over again with `SPC f e R`, pushes have been replaced by
add-to-list calls in the new macro).
All these steps are now put together at one place with the new macro
spacemacs|add-company-backends, check its docstring for more info on its
arguments.
This macro also allows to define arbitrary buffer local variables to tune
company for specific modes (similar to layer variables via a keyword :variables)
The code related to company backends management has been moved to the
auto-completion layer in the funcs.el file. A nice side effect of this move is
that it enforces correct encapsulation of company backends related code. We can
now easily detect if there is some configuration leakage when the
auto-completion layer is not used. But we loose macro expansion at file loading
time (not sue it is a big concern though).
The function spacemacs|enable-auto-complete was never used so it has been
deleted which led to the deletion of the now empty file core-auto-completion.el.
The example in LAYERS.org regarding auto-completion is now out of date and has
been deleted. An example to setup auto-completion is provided in the README.org
file of the auto-completion layer.
I find myself wanting to open the package.el and f3 is just too far away
in my keyboard. And I thought maybe this would be something people might
want. Also maybe we could have special binding for actions? like `C-1`
`C-2`, etc
- Apply `/` and `//` rules (double / is for private functions)
- Add missing `spacemacs/` prefixes
- Move functions used outside of spacemacs-base layer to
core/core-funcs.el
- Remove unused functions
Commit originally intented to only rename linum-update-window-scale-fix
to spacemacs/linum-update-window-scale-fix :-)
Fixes some edge cases like SPC f e R performed after SPC h SPC which
could wrongly install or uninstall packages.
Side effects is contained using the variable
configuration-layer--package-properties-read-onlyp, if non nil then
properties value of a package cannot be overwritten.
New value `manual` for auto-completion-enable-help-tooltip enables the
user to display help tooltip on-demand. Other values remain the same:
nil for no tooltip at all, t (or any non-nil non-manual value) for
automatic tooltip.
Minor fix: call `company-quickhelp-mode` just once, because it's a
global minor-mode.
Current implementation, #1644, doesn't handle the nested snippet
expansion properly, so you end up with smarparens disabled. Very
annoying.
I improved the implementation to remember the current state of
smartaperns before a top-level snippet expansion and preserve on any
nested snippet expansions.
Also, some housekeeping of the related code:
* Move hook handlers to the `funcs.el`.
* Add comments about yasnippet hooks subtleties relevant to the issue.
Fixes#1512
TL;DR Should get 20~25% speed improvement on startup, should get a big
improvement when using ivy or helm SPC h SPC. Users with layers.el files
in their layers must use `configuration-layer/declare-used-layer`
instead of `configuration-layer/declare-layer`
The implementation of the layer system made heavy use of `object-assoc`
and `object-assoc-list` functions which are not efficient. This PR
mainly replaces those object lists with hash maps in order to index the
objects by their name and achieve an O(1) access time.
The old object lists `configuration-layer--layers` and
`configuration-layer--packages` have been each by two variables each:
- `configuration-layer--indexed-layers` which is a hash-map of all the
layer objects and `configuration-layer--used-layers` which is a list of
all _used_ layers symbols,
- symmetrically `configuration-layer--indexed-packages` which is a
hash-map of all the package objects and
`configuration-layer--used-packages` which is a list of all _used_
packages symbols.
The hash map `configuration-layer--layer-paths` is gone, now we create
directly layer objects when discovering the layers and set the :dir
property. Note that previously the layer paths were the parent directory
of the layer, now :dir is the layer path.
The function `configuration-layer//make-layer` is now similar to its
counterpart `configuration-layer//make-package` in the sense that it
takes an optional `obj` to be able to override its properties.
The functions `configuration-layer/declare-layer` and
`configuration-layer/declare-layers` now takes an optional parameter
`usedp` in order to declare used or not used layers. For convenience
new functions have been added: `configuration-layer/declare-used-layer`
and `configuration-layer/declare-used-layers`, users _must_ update all
occurrences of `configuration-layer/declare-layer` by
`configuration-layer/declare-used-layers` in their `layers.el` files.
`helm-spacemacs-help` and `ivy-spacemacs-help` are updated to match the
changes in `core-configuration-layer.el`.
Rename some variables to make them more explicit:
`configuration-layer-no-layer` -> `configuration-layer-exclude-all-layers`
`configuration-layer-distribution` -> `configuration-layer-force-distribution`
With this new variable, user can load spacemacs anywhere, e.g.
"~/.emacs.d/spacemacs/". Only user's cache directory is still hard-coded
as "~/.emacs.d/.cache/". If user want to use spacemacs this way, drop
one line as the below in "~/.emacs.d/init.el":
(setq spacemacs-start-directory "~/.emacs.d/spacemacs/")
(load-file (concat spacemacs-start-directory "init.el"))
Implement a keymap for Spacemacs’ counsel search that supports saving
the results to a buffer (bound on F3 as in helm). Ensure that the new
buffer supports the GNE API.
As it was the case during the good old days when spacemacs was still a
baby, SPC h SPC now list correctly all the pre/init/post functions and
the associsted layer. It also now list properly the current owner of
a package so it is easy to find the owner of a package by narrowing
with `packagename owner`.
This new old functionality is possible by leveraging the last
improvements to the layer system (see previous commits).
TODO: update the sources for Ivy
Correctly augment the docstring based on the presence of b,f and /
bindings using dynamic hint.
b,f and / are not available if none of helm or ivy layers are used.
Invoking "make" or "project compilation" (<kbd>SPC c m</kbd>, <kbd>SPC c
c</kbd>) is currently not supported in ivy mode. The PR re-enables them
by adding helm-make to the ivy packages with ivy completion method.
Keep it in a recommendation in the README for the layer. Eval'ing the
custom-set-faces directly in the config makes it difficult to revert the
customizations.
Use a custom package name. Also dispatch functions to funcs.el files.
Disable ivy-hydra for now.
With this default configuration users using helm instead of ivy still
get a good ivy base configuration (UI and Key bindings).
Use a custom package name. Also dispatch functions to funcs.el files.
With this default configuration users using ivy instead of helm still
get a good helm base configuration (UI and Key bindings).
counsel-descbinds is accessible through SPC-?, and shadowing the C-h key
for which key makes pagination impossible. As this deviates from
which-key defaults, this should be up to user to config, not a base in
spacemacs-ivy.
Now Spacemacs will ask for layer installation when opening a file with
a known file type.
The auto-mode-alist entries are added by the file auto-layer.el in
the layers directory.
Easy insert of forms for lazy initialization can be done with the
interactive function configuration-layer//insert-lazy-install-form.
Change default value of dotspacemacs-enable-lazy-installation to t.
Motivation
While disabling Evil in holy-mode makes its implementation shorter and
sounds elegant on the paper, in practice it puts a big burden on the
configuration parts which need to know if Evil is enable or not. This is
a bad separation of concerns and the bunch of fixes that we were forced
to do in the past weeks shows this issue. Those fixes were about
removing the knowledge of the activation of Evil by implementing new
dispatching functions to be used by layers, this is cumbersome and makes
Spacemacs layer configuration more subtle which is not good. There was
additional bad consequences of the removal of Evil state like the
impossibility to use Evil lisp state or iedit states, or we would have
been forced to implement a temporary activation of Evil which is
awkward.
Instead I reintroduce Evil as the central piece of Spacemacs design thus
Evil is now re-enabled in holy-mode. It provides the abstraction we need
to isolate editing styles and be able to grow the Spacemacs
configuration coverage sanely. Layers don't need to check whether the
holy mode is active or not and they don't need to know if Evil is
available (it is always available). We also don't need to write
additional dispatching functions, this is the job of Evil, and I think
it provides everything for this. Ideally configuration layer should be
implemented with only Evil in mind and the holy-mode (and hybrid-mode)
should magically make it work for Emacs style users, for instance we can
freely use `evil-insert-state` anywhere in the code without any guard.
Evil is now even more part of Spacemacs, we can really say that
Spacemacs is Emacs+Evil which is now an indivisible pair. Spacemacs
needed this stable API to continue on the right track.
While these changes should be rather transparent to the user, I'm sorry
for this experimental period, I failed to see all the implications of
such a change, I was just excited about the possibility to make Evil
optional. The reality is that Spacemacs has to embrace it and keep its
strong position on being Emacs+Evil at the core.
Implementation
- insert, motion and normal states are forced to emacs state using an
advice on `evil-insert-state`, `evil-motion-state` and
`evil-normal-state` respectively. These functions can be used freely in
the layer configuration.
- A new general hook `spacemacs-editing-style-hook` allow to hook any
code that need to be configured based on the editing style. Functions
hooked to this hook takes the current style as parameter, this
basically generalize the hook used to setup hjkl navigation bindings.
- ESC has been removed from the emacs state map.
- Revert unneeded changes
- Revert "evil: enter insert-state only from normal-state"
commit bdd702dfbe.
- Revert "avoid being evil in deft with emacs editing style"
commit f3a16f49ed.
Additional changes
All editing style packages have been moved to a layer called
`spacemacs-editing-styles`
Notes
I did not have time to attack hybrid mode, I should be able to do it
later.