Language Servers may signal that they do not support a method in
the initialization result (server capabilities). We can check these
when making LSP requests and hint in the status line when a method
is not supported by the server. This can also prevent crashes in
servers which assume that clients do not send requests for methods
which are disabled in the server capabilities.
There is an existing pattern the LSP client module where a method
returns `Option<impl Future<Output = Result<_>>>` with `None` signaling
no support in the server. This change extends this pattern to the rest
of the client functions. And we log an error to the statusline for
manually triggered LSP calls which return `None`.
Previously, jumplists could grow unchecked. Every transaction is
applied to jumplist selections to ensure that they are up to date
and within document bounds, so this would cause every edit to become
more expensive as jumplist lengths increased throughout a session.
Setting a maximum number of entries limits the cost.
Vim and Neovim limit their jumplists:
* b298fe6cba/src/structs.h (L141)
* e8cc489acc/src/nvim/mark_defs.h (L57)
Notably, Kakoune does not. In Kakoune, changes are applied to jumplist
entries lazily as you hit `<C-o>`/`<C-i>` though, so Kakoune doesn't
have the same growing cost concerns. Kakoune also does not have a
concept of a View which limits the cost further.
Vim and Neovim limit to 100. This seems unreasonably high to me so I've
set this to 30 to start. We can increase if this is problematically
low.
d6323b7cbc changed the behavior of paste
to select the newly inserted text. This is preferrable in normal mode
because it's useful to be able to act on the new text. This behavior
is worse for insert or select mode though:
* In insert mode, the cursor ends up on the last character of the newly
selected text, so further typing inserts text before the last
character.
* In select mode, the current selection is replaced with the new text
selection which doesn't extend the current selection. With this
change, the selection is extended to include the new text.
This aligns the behavior more closely with Kakoune, but it's
coincidental instead of intentional: Kakoune doesn't implement
bracketed paste (AFAIK) which causes this behavior in insert mode,
and Kakoune doesn't have a select mode.
Previously, commands such as `r<tab>` (replace with tab) or `t<tab>`
(select till tab) had no effect. This is because `KeyCode::Tab` needs
special treatment (like `KeyCode::Enter`).
This change handles a language server exiting. This was a UX sore-spot:
if a language server crashed, Helix did not recognize the exit and
continued to send requests to it. All requests would timeout since they
would not receive responses. This would also hold-up Helix closing
itself down since it would try to gracefully shutdown the server which
is implemented in the LSP spec as a request.
We could attempt to automatically restart the language server on crash.
I left this for future work since that change will need to be slightly
complicated: it will need to cover the case of a language server
repeatedly crashing.
d7d0d5ffb7 resolves completion items on
the idle-timeout event. The `Completion::resolve_completion_item`
function blocks on the LSP request though, which blocks the compositor
and in turn blocks the event loop. So until the language server returns
the resolved completion item, Helix is unable to respond to keypresses
or other LSP messages.
This is typically ok since the resolution request is fast but for some
language servers this can be problematic, and ideally we shouldn't be
blocking like this anyways.
When receiving a `completionItem/resolve` request, the Volar server
sends a `workspace/configuration` request to Helix and blocks itself
on the response, leading to a deadlock. Eventually the resolve request
times out within Helix but Helix is locked up and unresponsive in that
window.
This change resolves the completion item without blocking the
compositor.
PR #4134 switched the autocomplete menu from alphabetical to fuzzy
sorting. This commit removes the still existing filtering by prefix and
should enable full fuzzy sorting of the autocomplete menu.
closes#3084, #1807
Co-authored-by: Blaž Hrastnik <blaz@mxxn.io>
This fixes an edge case for completing shellwords. With a file
"a b.txt" in the current directory, the sequence `:open a\<tab>`
will result in the prompt containing `:open aa\ b.txt`. This is
because the length of the input which is trimmed when replacing with
completion is calculated on the part of the input which is parsed by
shellwords and then escaped (in a separate operation), which is lossy.
In this case it loses the trailing backslash.
The fix provided here refactors shellwords to track both the _words_
(shellwords with quotes and escapes resolved) and the _parts_ (chunks
of the input which turned into each word, with separating whitespace
removed). When calculating how much of the input to delete when
replacing with the completion item, we now use the length of the last
part.
This also allows us to eliminate the duplicate work done in the
`ends_with_whitespace` check.
The text within the command palette used a custom format to display
the keybinding for a command. This change switches to the key sequence
format that we use for pending keys and macros.
* init
* cargo fmt
* optimisation of the scrollbar render both for Menu and Popup. Toggling off scrollbar for Popup<Menu>, since Menu has its own
* rendering scroll track
* removed unnecessary cast
* improve memory allocation
* small correction
d6323b7cbc introduced a regression for
shell commands like `|`, `!`, and `<A-!>` which caused the new
selections to be incorrect. This caused a panic when piping (`|`)
would cause the new range to extend past the document end.
The paste version of this bug was fixed in
48a3965ab4.
This change also inherits the direction of the new range from the old
range and adds integration tests to ensure that the behavior isn't
broken in the future.
* dynamically resize line number gutter width
* removing digits lower-bound, permitting spacer
* removing max line num char limit; adding notes; qualified successors; notes
* updating tests to use new line number width when testing views
* linenr width based on document line count
* using min width of 2 so line numbers relative is useful
* lint rolling; removing unnecessary type parameter lifetime
* merge change resolution
* reformat code
* rename row_styler to style; add int_log resource
* adding spacer to gutters default; updating book config entry
* adding view.inner_height(), swap for loop for iterator
* reverting change of current! to view! now that doc is not needed
If `a\ b.txt` were a local file, `:o a\ <tab>` would fill the prompt
with `:o aa\ b.txt` because the replacement range was calculated using
the shellwords-parsed part. Escaping the part before calculating its
length fixes this edge-case.
This changes the completion items to be rendered with shellword
escaping, so a file `a b.txt` is rendered as `a\ b.txt` which matches
how it should be inputted.
8584b38cfb switched to shellwords for
completion in command-mode. This changes the conditions for choosing
whether to complete the command or use the command's completer.
This change processes the input as shellwords up-front and uses
shellword logic about whitespace to determine whether the command
or argument should be completed.
* Fix range offsets in multi-selection paste
d6323b7cbc introduced a regression with
multi-selection paste where pasting would not adjust the ranges
correctly. To fix it, we need to track the total number of characters
inserted in each changed selection and use that offset to slide each
new range forwards.
* Inherit selection directions on paste
* Add an integration-test for multi-selection pasting
The sequence "_y"_p panics because the blackhole register contains an
empty values vec. This causes a panic when pasting since it unwraps
a `slice::last`.
This follows changes in Kakoune to the same effects:
* p/<space>p: 266d1c37d0
* !/<A-!>: 85b78dda2e
Selecting the new data inserted by shell or pasting is often more
useful than retaining a selection of the pre-paste/insert content.
* Clamp highlighting range to be within document
This fixes a panic possible when two vsplits of the same document
exist and enough lines are deleted from the document so that one of
the windows focuses past the end of the document.
* Ensure cursor is in view on window change
If two windows are editing the same document, one may delete enough of
the document so that the other window is pointing at a blank page (past
the document end). In this change we ensure that the cursor is within
view whenever we switch to a new window (for example with `<C-w>w`).
* Update helix-term/src/ui/editor.rs
Co-authored-by: Blaž Hrastnik <blaz@mxxn.io>
Co-authored-by: Blaž Hrastnik <blaz@mxxn.io>
When backward-deleting a character, if this character and the following
character form a Pair, we want to delete both. However, there is a bug
that deletes both characters also if both characters are closers of some
Pair.
This commit fixes that by adding an additional check that the deleted
character should be an opener in a Pair.
Closes https://github.com/helix-editor/helix/issues/4544.
Most commands that accept an argument show their current value if no
argument is specified. The `:theme` command previously displayed an
error message in the status bar if not provided with an argument:
```
Theme name not provided
```
It now shows the current theme name in the status bar if no argument is
specified.
Signed-off-by: James O. D. Hunt <jamesodhunt@gmail.com>
Signed-off-by: James O. D. Hunt <jamesodhunt@gmail.com>
This bug occurs on `shell_insert_output` and `shell_append_output`
commands.
The previous implementation would create a child process using the Rust
stdlib's `Command` builder. However, when nothing should be piped in
from the editor, the default value for `stdin` would be used. According
to the Rust stdlib documentation that is `Stdio::inherit` which will
make the child process inherit the parent process' stdin. This would
cause the terminal to freeze.
This change will set the child process' stdin to `Stdio::null` whenever
it doesn't pipe it. In the `if` statement where this change was made
there was an extra condition for windows that I am not sure if would
require some special treatment.
This is mostly for the sake of the diagnostics pickers: without
rendering the diagnostic styles, it's hard to tell where the entries
in the picker are pointing to.
Some language servers may not send the `documentation` field if it
is expensive to compute. Clients can request the missing field with
a completionItem/resolve request.
In this change we use the idle-timeout event to ensure that the current
completion item is resolved.
This complicates the code a little but it often divides by two the number of allocations done by
the functions. LSP labels especially can easily be called dozens of time in a single menu popup,
when listing references for example.
When we do auto formatting, the code that takes the LSP's response and applies
the changes to the document are just getting the currently focused view and
giving that to the function, basically always assuming that the document that
we're applying the change to is in focus, and not in a background view.
This is usually fine for a single view, even if it's a buffer in the
background, because it's still the same view and the selection will get updated
accordingly for when you switch back to it. But it's obviously a problem for
when there are multiple views, because if you don't have the target document in
focus, it will ask the document to update the wrong view, hence the crash.
The problem with this is picking which view to apply any selection change to.
In the absence of any more data points on the views themselves, we simply pick
the first view associated with the document we are saving.
When force quitting, we need to block on the pending writes to ensure
that write commands succeed before exiting, and also to avoid a crash
when all the views are gone before the auto format call returns from
the LS.
* Autosave all when the terminal loses focus
* Correct comment on focus config
Co-authored-by: Blaž Hrastnik <blaz@mxxn.io>
* Need a block_try_flush_writes in all quit_all paths
Co-authored-by: Blaž Hrastnik <blaz@mxxn.io>
* Fix test::print for Unicode
The print function was not generating correct translations when
the input has Unicode (non-ASCII) in it. This is due to its use of
String::len, which gives the length in bytes, not chars.
* Fix multi-code point auto pairs
The current code for auto pairs is counting offsets by summing the
length of the open and closing chars with char::len_utf8. Unfortunately,
this gives back bytes, and the offset needs to be in chars.
Additionally, it was discovered that there was a preexisting bug where
the selection was not computed correctly in the case that the cursor
was:
1. a single grapheme in width
2. this grapheme was more than one char
3. the direction of the cursor is backwards
4. a secondary range
In this case, the offset was not being added into the anchor. This was
fixed.
* migrate auto pairs tests to integration
* review comments
This change removes language server configuration from the default
languages.toml config for integration tests. No integration-tests
currently depend on the availability of a language server but if any
future test needs to, it may provide a language server configuration
by passing an override into the `test_syntax_conf` helper.
Language-servers in integration tests cause false-positive failures
when running integration tests in GitHub Actions CI. The Windows
runner appears to have `clangd` installed and all OS runners have
the `R` binary installed but not the `R` language server package.
If a test file created by `tempfile::NamedTempFile` happens to have a
file extension of `r`, the test will most likely fail because the
R language server will fail to start and will become a broken pipe,
meaning that it will fail to shutdown within the timeout, causing a
false-positive failure. This happens surprisingly often in practice.
Language servers (especially rust-analyzer) also emit unnecessary
log output when initializing, which this change silences.
`helix_view::apply_transaction` closes over `Document::apply` and
`View::apply` to ensure that jumplist entries are updated when a
document changes from a transaction. `Document::apply` shouldn't
be called directly - this helper function should be used instead.
If a document is written with a new path, currently, in the event that
the write fails, the document still gets its path changed. This fixes
it so that the path is not updated unless the write succeeds.
The way that document writes are handled are by submitting them to the
async job pool, which are all executed opportunistically out of order. It
was discovered that this can lead to write inconsistencies when there
are multiple writes to the same file in quick succession.
This seeks to fix this problem by removing document writes from the
general pool of jobs and into its own specialized event. Now when a
user submits a write with one of the write commands, a request is simply
queued up in a new mpsc channel that each Document makes to handle its own
writes. This way, if multiple writes are submitted on the same document,
they are executed in order, while still allowing concurrent writes for
different documents.
Instead of repeatedly checking if it is in_bounds, calculate the
max_indent beforehand and just loop. I added a debug_assert to "prove"
that it never tries drawing out of bounds.
Better performance, and otherwise very long lines with lots of tabs
will wrap around the u16 and come back on the other side, messing up
the beginning skip_levels.
Also changes workspace diagnostic picker bindings to <space>D and
changes the debug menu keybind to <space>g, the previous diagnostic
picker keybind. This brings the diagnostic picker bindings more in
line with the jump to next/previous diagnostic bindings which are
currently on ]d and [d.
The debug assertion that document diagnostics are sorted incorrectly
panics for cases like `[161..164, 162..162]`. The merging behavior
in the following lines that relies on the assertion only needs the
input ranges to be sorted by `range.start`, so this change simplifies
the assertion to only catch violations of that assumption.
Undo/redo/earlier/later call `Document::apply_impl` which applies
transactions to the document. These transactions also need to be
applied to the view as in 0aedef0.
Here we separate the diagnostics by severity and then overlay the Vec
of spans for each severity on top of the highlights. The error
diagnostics end up overlaid on the warning diagnostics, which are
overlaid on the hints, overlaid on info, overlaid on any other severity
(default), then overlaid on the syntax highlights.
This fixes two things:
* Error diagnostics are now always visible when overlapped with other
diagnostics.
* Ghost text is eliminated.
* Ghost text was caused by duplicate diagnostics at the EOF:
overlaps within the merged `Vec<(usize, Range<usize>)>` violate
assumptions in `helix_core::syntax::Merge`.
* When we push a new range, we check it against the last range and
merge the two if they overlap. This is safe because they both
have the same severity and therefore highlight.
The actual merge is skipped for any of these when they are empty, so
this is very fast in practice. For some data, I threw together an FPS
counter which renders as fast as possible and logs the renders per
second.
With no diagnostics, I see an FPS gain from this change from 868 FPS
to 878 (+1.1%) on a release build on a Rust file. On an Erlang file
with 12 error diagnostics and 6 warnings in view (233 errors and 66
warnings total), I see a decrease in average FPS from 795 to 790
(-0.6%) on a release build.
It is easy to forget to call `Document::apply` and/or `View::apply` in
the correct order. This commit introduces a helper function which
closes over both calls.
This change adds View::apply calls for all Document::apply call-sites,
ensuring that changes to a document do not leave invalid entries in
the View's jumplist.
* Implement cursorcolumn
* Add documentation
* Separate column style from line with fallback
* Fallback to cursorcolumn first
* Switch to non-fallback try_get_exact
Add new function `try_get_exact`, which doesn't perform fallback,
and use that instead because the fallback behaviour is being handled
manually.
If the close method fails, the editor will quit before restoring the
terminal. This causes the shell to break if, e.g. the LS times out
shutting down.
This fixes this by always restoring the terminal after closing, and
printing out a message to stderr if there is an error.
This change automatically tracks pending text for for commands which use
on-next-key callbacks. For example, `t` will await the next key event
and "t" will be shown in the bottom right-hand corner to show that we're
in a pending state.
Previously, the text for these on-next-key commands needed to be
hard-coded into the command definition which had some drawbacks:
* It was easy to forget to write and clear the pending text.
* If a command was remapped in a custom config, the pending text would
still show the old key.
With this change, pending text is automatically tracked based on the
key events that lead to the command being executed. This works even
when the command is remapped in config and when the on-next-key
callback is nested under some key sequence (for example `mi`).
* Keep arrow and special keys in insert
Advanced users won't need it and is useful for beginners.
Revert part of #3671.
* Change text for insert mode section
Co-authored-by: Blaž Hrastnik <blaz@mxxn.io>
* Remove ctrl-up/down in insert
* Reorganize insert keys and docs
* Improve page up experience on last tutor
The last tutor page can page down multiple times and it will break the
heading on the 80x24 screen paging when reaching the last page, this
keeps the style the same and make sure page up and down won't break it.
Co-authored-by: Blaž Hrastnik <blaz@mxxn.io>
* Change focus to modified docs on quit
When quitting with modified documents, automatically switch focus to
one of them.
* Update helix-term/src/commands/typed.rs
Co-authored-by: Poliorcetics <poliorcetics@users.noreply.github.com>
* Make it work with buffer-close-all and the like
* Cleanup
Use Cow instead of String, and rename DoesntExist -> DoesNotExist
Co-authored-by: Poliorcetics <poliorcetics@users.noreply.github.com>
* Add option to skip the first indent guide
* reorder skip_first option
* change indent-guides.skip_first to a number
* rename skip -> skip_levels
* add skip_levels to the book
* Update book/src/configuration.md
Co-authored-by: A-Walrus <58790821+A-Walrus@users.noreply.github.com>
* Update helix-term/src/ui/editor.rs
Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
Co-authored-by: Robin <robinvandijk@klippa.com>
Co-authored-by: A-Walrus <58790821+A-Walrus@users.noreply.github.com>
Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
* keymap: Rename A "Insert at end of line"
The language for the `A` binding is potentially confusing because
`A` behaves like `i` done at the end of the line rather than `a`.
This change renames the command to match Kakoune's language[^1].
[^1]: 021da117cf/src/normal.cc (L2229)
* keymap: Rename I `insert_at_line_start`
* Select inserted space after join
* Split join_selections with space selection to A-J
Kakoune does that too and some users may still want to retain their selections.
* Update join_selections docs
When signature help is too large it may cause a panic when it is too
large, now I just make the hover do an intersection with surface to make
sure it never overflow.
This changes the behavior of operations like `]f`/`[f` to set the
direction of the new range to the direction of the action.
The original behavior was to always use the head of the next function.
This is inconsistent with the behavior of goto_next_paragraph and makes
it impossible to create extend variants of the textobject motions.
This causes a behavior change when there are nested functions. The
behavior in the parent commit is that repeated uses of `]f` will
select every function in the file even if nested. With this commit,
functions are skipped.
It's notable that it's possible to emulate the original behavior by
using the `ensure_selections_forward` (A-:) command between invocations
of `]f`.
* Show "Invalid regex" message on enter (Validate)
* Reset selection on invalid regex
* Add popup for invalid regex
* Replace set_position with position
* Make popup auto close
* Split helix_core::find_root and helix_loader::find_local_config_dirs
The documentation of find_root described the following priority for
detecting a project root:
- Top-most folder containing a root marker in current git repository
- Git repository root if no marker detected
- Top-most folder containing a root marker if not git repository detected
- Current working directory as fallback
The commit contained in https://github.com/helix-editor/helix/pull/1249
extracted and changed the implementation of find_root in find_root_impl,
actually reversing its result order (since that is the order that made
sense for the local configuration merge, from innermost to outermost
ancestors).
Since the two uses of find_root_impl have different requirements (and
it's not a matter of reversing the order of results since, e.g., the top
repository dir should be used by find_root only if there's not marker in
other dirs), this PR splits the two implementations in two different
specialized functions.
In doing so, find_root_impl is removed and the implementation is moved
back in find_root, moving it closer to the documented behaviour thus
making it easier to verify it's actually correct
* helix-core: remove Option from find_root return type
It always returns some result, so Option is not needed
* Improve keymap errors from command typos
Currently, opening helix with a config containing a bad command mapping
fails with a cryptic error. For example, say we have a config (bad.toml)
with a command name that doesn't exist:
[keys.normal]
b = "buffer_close" # should be ":buffer-close"
When we `hx -c bad.toml`, we get...
> Bad config: data did not match any variant of untagged enum KeyTrie for key `keys.normal` at line 1 column 1
> Press <ENTER> to continue with default config
This is because of the way that Serde tries to deserialize untagged
enums such as `helix_term::keymap::KeyTrie`. From the Serde docs[^1]:
> Serde will try to match the data against each variant in order and the
> first one that deserializes successfully is the one returned.
`MappableCommand::deserialize` fails (returns an Err variant) when a
command does not exist. Serde interprets this as the `KeyTrie::Leaf`
variant failing to match and declares that the input data doesn't
"match any variant of untagged enum KeyTrie."
Luckily the variants of KeyTrie are orthogonal in structure: we can tell
them apart by the type hints from a `serde:🇩🇪:Visitor`. This change
uses a custom Deserialize implementation along with a Visitor that
discerns which variant of the KeyTrie applies. With this change, the
above failure becomes:
> Bad config: No command named 'buffer_close' for key `keys.normal.b` at line 2 column 5
> Press <ENTER> to continue with default config
We also provide more explicit information about the expectations on
the field. A config with an unexpected type produces a message with
that information and the expectation:
[keys.normal]
b = 1
> Bad config: invalid type: integer `1`, expected a command, list of commands, or sub-keymap for key `keys.normal.b` at line 2 column 5
> Press <ENTER> to continue with default config
[^1]: https://serde.rs/enum-representations.html#untagged
* Update helix-term/src/keymap.rs
Co-authored-by: Ivan Tham <pickfire@riseup.net>
Co-authored-by: Ivan Tham <pickfire@riseup.net>
* Add command line parameter to specify log file
I had the logs of my debug helix mixed in with the logs from the
production helix.
Add a `--log` command line argument to redirect any logs to other
files, making my debugging easier :-)
* Update completion files with `--log` argument
The tutor file is loaded as .txt which can potentially spawn a
language server. Then the path is unset, but the LS remains active.
This can cause panics since updates are now submitted for a doc
with no path.
As a quick workaround we remove the extension which should avoid
detection.
Fixes#3730
* Don't change config to default when refreshing invalid config
* Propely handle theme errors with config-reload
* Extract refresh theme into seperate function
Helix is first and foremost a modal editor. Willingness to support non-modal
editing is there, but it is not one that should be encouraged with the default
settings. There are an increasing number of users who are stumbling because
they are trying to use Helix as a non-modal editor, so this is an effort to
encourage new users to stop and take notice that Helix has a different paradigm
than VSCode, Sublime, etc. Users can still add these bindings back to their own
configs if they wish.
`extend_line_above` (and `extend_line` when facing backwards) skip
a line when the current range does not fully cover a line.
Before this change:
foo
b#[|a]#r
baz
With `extend_line_above` or `extend_line` selected the line above.
#[|foo
bar]#
baz
Which is inconsistent with `extend_line_below`. This commit changes
the behavior to select the current line when it is not already
selected.
foo
#[|bar]#
baz
Then further calls of `extend_line_above` extend the selection up
line-wise.
* fix: Recalculate completion when going through prompt history
* Update completion when the prompt line is changed
It should not be possible to update the line without also updating the
completion since the completion holds an index into the line.
* Fix Prompt::with_line recalculate completion
with_line was the last function where recalculate completion had to be
done manually. This function now also recalculates the completion so
that it's impossible to forget.
* Exit selection when recalculating completion
Keeping the selection index when the completion has been recalculated
doesn't make sense. This clears the selection automatically, removing
most needs to manually clear it.
* Remove &mut on save_filter
Co-authored-by: Blaž Hrastnik <blaz@mxxn.io>
It was starting to diverge as the normal exit code was restoring the prompt but the panic code
wasn't, and the panic code was disabling bracketed paste but the normal code wasn't.
This changes the panic path slightly in that we won't disable raw mode if exiting alternate screen
and disabling bracketed paste fails. If that happens, things are so busted I don't think it matters
anyway.
Fixes a panic with a config like:
[keys.normal.space]
x = [":buffer-close"]
by bailing out of the command-execution handling if the document
doesn't exist after handling a command.
This refactor changes the overall structure of the goto_ts_object_impl
command without removing any functionality from its behavior. The
refactored motion:
* acts on all selections instead of reducing to one selection
* may be repeated with the `repeat_last_motion` (A-.) command
* informs the user when the syntax-tree is not accessible in the current buffer
This is invalid according to the [LSP spec]:
> In addition the server is not allowed to send any requests
> or notifications to the client until it has responded with an
> InitializeResult, with the exception that during the initialize
> request the server is allowed to send the notifications
> window/showMessage, window/logMessage and telemetry/event as well
> as the window/showMessageRequest request to the client.
So we should discard the message when the language server is not
yet initialized. This can happen if the server sends
textDocument/publishDiagnostics before responding to the initialize
request. clojure-lsp appears to exhibit this behavior in the wild.
[LSP Spec]: https://microsoft.github.io/language-server-protocol/specifications/lsp/3.17/specification/#initialize
This change adds documents to the view's document history Vec.
(This is used by `ga` for example to access the last buffer.)
Previously, a sequence like so would have confusing behavior:
1. Open file A: any document with an active language server
2. Find some definition that lives in another file - file B - with `gd`
3. Jump back in the jumplist with `C-o` to file A
4. Use `ga` intending to switch back to file B
The behavior prior to this change was that `ga` would switch to file
A: you could not use `ga` to switch to file B.
When changing focus, the lookup with `current!` may change the
view and end up executing mode transition hooks on the newly
focused view. We should use the same view and document to execute
mode transition hooks so that switching away from a view triggers
history save points.
* Derive Document language name from `languages.toml` `name` key
This changes switches from deriving the language name from the
`languages.toml` `scope` key to `name` (`language_id` in the
`LanguageConfiguration` type). For the most part it works to derive the
language name from scope by chopping off `source.` or `rsplit_once` on
`.` but for some languages we have now like html (`text.html.basic`),
it doesn't. This also should be a more accurate fallback for the
`language_id` method which is used in LSP and currently uses the
`rsplit_once` strategy.
Here we expose the language's name as `language_name` on `Document` and
replace ad-hoc calculations of the language name with the new method.
This is most impactful for the `file-type` statusline element which is
using `language_id`.
* Use `Document::language_name` for the `file-type` statusline element
The `file-type` indicator element in the statusline was using
`Document::language_id` which is meant to be used to for telling
Language Servers what language we're using. That works for languages
with `language-server` configurations in `languages.toml` but shows
text otherwise. The new `Document::language_name` method from the
parent commit is a more accurate way to determine the language.
* let extend-line respect range direction
* fix extend above logic
* keep `x` existing binding
* Update book/src/keymap.md
Co-authored-by: Ivan Tham <pickfire@riseup.net>
Co-authored-by: Ivan Tham <pickfire@riseup.net>
* Update description of `m` textobject to its actual functionality
Sometime recently the functionality of `m` was changed to match the
nearest pair to the cursor, rather than the former functionality of
matching the pair only if the cursor was on one of the brace characters
directly.
* Rename surround methods to reflect that they work on pairs
The current naming suggests that they may work generally on any
textobject, whereas their implementation really focuses on pairs.
* Change description of m textobject to match actual functionality
The current implementation of `m` no longer merely looks at the pair
character the cursor is on, but actually will search for the pair
(defined in helix-core/src/surround.rs) that encloses the cursor, and
not the entire selection.
* Accept suggested wording change
Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
* Prefix pair surround for consistency
Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
Ported over from 61365dfbf3 in the `gui` branch. This will allow
adding our own events, most notably an idle timer event (useful
for adding debounced input in [dynamic pickers][1] used by interactive
global search and workspace symbols).
[1]: https://github.com/helix-editor/helix/pull/3110
Co-authored-by: Blaž Hrastnik <blaz@mxxn.io>
These are read-line-like bindings which we'd like to minimize in
insert mode in general.
In particular these two are troublesome if you have a low
`editor.idle-timeout` config and are using LSP completions: the
behavior of C-n/C-p switches from moving down/up lines to moving
down/up the completion menu, so if you hit C-n too quickly
expecting to be in the completion menu, you'll end up moving down
a line instead. Using C-p moves you back up the line but doesn't
re-trigger the completion menu. This kind of timing related change
to behavior isn't realistically that big of a deal but it can be
annoying.
* Fix tab highlight when tab is partially visible
* Make it style based, and not truncation based
Dealing with truncating is a mess, especially when it comes to wide
unicode graphemes. This way it should work no matter what.
* Inline style calculation into branches
* Fix incorrect indent guide styling
Before the indent guides on top of whitespace inherited the theme
from them. Now they do not.
* Fix dark_plus theme indent_guides
* Use whitespace style as fallback for indent-guide
* Fix dark_plus theme indent_guides
* Move indent_guide style patching out of loop
* Avoid setting stdin handle when not necessary
Avoid setting the stdin handle in `shell_impl` when the input argument
is None.
This permits to run commands with no stdin with :sh
* refactoring to avoid code duplication
* making clippy happy
* Process variable name fix
Indent style may change when choosing a language with `:set-language`.
Line-endings most likely will not change, but `:set-language` should
have a similar effect as reloading a file (`:reload`), plus the two
are currently grouped in the implementation and line-ending detection
is not particularly expensive.
* Change default formatter for any language
* Fix clippy error
* Close stdin for Stdio formatters
* Better indentation and pattern matching
* Return Result<Option<...>> for fn format instead of Option
* Remove unwrap for stdin
* Handle FormatterErrors instead of Result<Option<...>>
* Use Transaction instead of LspFormatting
* Use Transaction directly in Document::format
* Perform stdin type formatting asynchronously
* Rename formatter.type values to kebab-case
* Debug format for displaying io::ErrorKind (msrv fix)
* Solve conflict?
* Use only stdio type formatters
* Remove FormatterType enum
* Remove old comment
* Check if the formatter exited correctly
* Add formatter configuration to the book
* Avoid allocations when writing to stdin and formatting errors
* Remove unused import
Co-authored-by: Gokul Soumya <gokulps15@gmail.com>
The language server sends a char offset range within the
signature help label text to highlight as the current parameter,
but helix uses byte offset ranges for rendering highlights. This
was brought up in the [review of the original signature help PR][1],
but the ranges were being highlighted correctly, and there were no
out of bound or indexing panics. Turns out rust-analyzer was
[incorrectly sending byte offsets] instead of char offsets and this
made it seem like all was well and good with offsets in helix during
initial testing.
[1]: https://github.com/helix-editor/helix/pull/1755#discussion_r906715371
[2]: https://github.com/rust-lang/rust-analyzer/pull/12272
* auto pair-removal
Fixes https://github.com/helix-editor/helix/issues/1673
* autopairs removal: use doc autopairs
* autopairs-removal: limit to one-char selections
* use single_grapheme() to check if range is one char
* fix errouneous deletes of " and other symmetric autopairs when at buffer start
Co-authored-by: Houkime <>
* add statusline element to display file line endings
* run cargo fmt --all
* change the word *ending* from plural to singular
* support for the unicode-lines feature flag
* Add lsp signature help
* Do not move signature help popup on multiple triggers
* Highlight current parameter in signature help
* Auto close signature help
* Position signature help above to not block completion
* Update signature help on backspace/insert mode delete
* Add lsp.auto-signature-help config option
* Add serde default annotation for LspConfig
* Show LSP inactive message only if signature help is invoked manually
* Do not assume valid signature help response from LSP
Malformed LSP responses are common, and these should not crash the
editor.
* Check signature help capability before sending request
* Reuse Open enum for PositionBias in popup
* Close signature popup and exit insert mode on escape
* Add config to control signature help docs display
* Use new Margin api in signature help
* Invoke signature help on changing to insert mode
* feat(statusline): add the file type (language id) to the status line
* refactor(statusline): move the statusline implementation into an own struct
* refactor(statusline): split the statusline implementation into different functions
* refactor(statusline): Append elements using a consistent API
This is a preparation for the configurability which is about to be
implemented.
* refactor(statusline): implement render_diagnostics()
This avoid cluttering the render() function and will simplify
configurability.
* feat(statusline): make the status line configurable
* refactor(statusline): make clippy happy
* refactor(statusline): avoid intermediate StatusLineObject
Use a more functional approach to obtain render functions and write to
the buffers, and avoid an intermediate StatusLineElement object.
* fix(statusline): avoid rendering the left elements twice
* refactor(statusline): make clippy happy again
* refactor(statusline): rename `buffer` into `parts`
* refactor(statusline): ensure the match is exhaustive
* fix(statusline): avoid an overflow when calculating the maximal center width
* chore(statusline): Describe the statusline configurability in the book
* chore(statusline): Correct and add documentation
* refactor(statusline): refactor some code following the code review
Avoid very small helper functions for the diagnositcs and inline them
instead.
Rename the config field `status_line` to `statusline` to remain
consistent with `bufferline`.
* chore(statusline): adjust documentation following the config field refactoring
* revert(statusline): revert regression introduced by c0a1870
* chore(statusline): slight adjustment in the configuration documentation
* feat(statusline): integrate changes from #2676 after rebasing
* refactor(statusline): remove the StatusLine struct
Because none of the functions need `Self` and all of them are in an own
file, there is no explicit need for the struct.
* fix(statusline): restore the configurability of color modes
The configuration was ignored after reintegrating the changes of #2676
in 8d28f95.
* fix(statusline): remove the spinner padding
* refactor(statusline): remove unnecessary format!()
Ctrl-based shortcuts are common in numerous applications.
This change:
- Adds Ctrl+{Left/Right/Backspace/Delete} for word-wise movement/deletion in prompt, picker, …
- Removes Alt-Left and Alt-Right in prompt, picker, …
- Adds Alt-Delete in insert mode for forward word deletion
In some terminals, Alt-Backspace might not work because it is ambigous.
See: https://github.com/helix-editor/helix/pull/2193#issuecomment-1105042501
Hence, Alt alternative is not removed.
* Fix backwards selection duplication widening bug
* Add integration tests
* Make tests line-ending agnostic
Make tests line-ending agnostic
Use indoc to fix tests
Fix line-ending on test input
* Refactor menu::Item to accomodate external state
Will be useful for storing editor state when reused by pickers.
* Add some type aliases for readability
* Reuse menu::Item trait in picker
This opens the way for merging the menu and picker code in the
future, since a picker is essentially a menu + prompt. More
excitingly, this change will also allow aligning items in the
picker, which would be useful (for example) in the command palette
for aligning the descriptions to the left and the keybinds to
the right in two separate columns.
The item formatting of each picker has been kept as is, even though
there is room for improvement now that we can format the data into
columns, since that is better tackled in a separate PR.
* Rename menu::Item::EditorData to Data
* Call and inline filter_text() in sort_text() completion
* Rename diagnostic picker's Item::Data
* Sort themes by score & then name
Previously the themes were appearing unordered after typing ':theme '.
This sorts them first by fuzzy score and then by name so that they
generally appear in a more ordered fashion in the initial list.
The sort by name does not really pay off when there is a score so an
alternative approach would be to sort by name if there is string to
fuzzy match against and otherwise sort by score.
I've lowercased the names as that avoids lower case & upper case letters
being sorted into separate groups. There might be a preferable approach
to that though.
* Sort language & files by score then name
And change to use sort_unstable_by instead of sort_unstable_by_key as it
allows us to avoid some allocations.
I don't fully understand the flow of the 'filename_impl' function but
this seems to deliver the desired results.
* Remove unnecessary reference
Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
* adds --vsplit and --hsplit arguments
* moved comment
* fixed lint (third time's a charm)
* changed vsplit and hsplit from two separate bools to type Option<Layout>, and some cleanup
* delete_backwards_char accepts any type of whitespace
* Fix inconsistency, where unicode whitespaces are treated as normal whitespaces
* Changed back to direct whitespace match
* Only accept explicit whitespace / tabs
Co-authored-by: s0LA1337 <dreamer@neoncity.dev>
* Add mode specific styles
In similar vein to neovim's lualine and similar statusline packages this
allows helix users to style their mode based on which mode it is thus
making each mode more visually distinct at a glance
* Add an example based on rosepine
* Add editor.colors-mode config
* Document statusline mode styles
* Add workspace and document diagnostics picker
fixes#1891
* Fix some of @archseer's annotations
* Add From<&Spans> impl for String
* More descriptive parameter names.
* Adding From<Cow<str>> impls for Span and Spans
* Add new keymap entries to docs
* Avoid some clones
* Fix api change
* Update helix-term/src/application.rs
Co-authored-by: Bjorn Ove Hay Andersen <bjrnove@gmail.com>
* Fix a clippy hint
* Sort diagnostics first by URL and then by severity.
* Sort diagnostics first by URL and then by severity.
* Ignore missing lsp severity entries
* Add truncated filepath
* Typo
* Strip cwd from paths and use url-path without schema
* Make tests a doctest
* Better variable names
Co-authored-by: Falco Hirschenberger <falco.hirschenberger@itwm.fraunhofer.de>
Co-authored-by: Bjorn Ove Hay Andersen <bjrnove@gmail.com>
- Simplified match statements by destructuring MouseEvent struct
at the top and then matching on event.kind.
- Extracted out closures for calculating (1) position and view
of mouse click and (2) gutter coordinates and view of mouse click.
* Add single width left margin for completion popup
* Clear with ui.menu style before rendering menu
When rendering a completion popup, the popup component will clear
the area with ui.popup and then the menu component would draw over
it using a table component. We remove the left edge of the area
before passing it to the table component (so that it will be left
as padding), and the table component uses ui.menu as the style.
If ui.menu and ui.popup are different the left edge of the popup
will look different from the rest of the popup. We avoid this by
clearing the whole area with ui.menu in Menu::render
* check selection's visible width when copying on mouse click
Mouse-click-up copies the selection produced by dragging. The event
is ignored if the selection has a width of 1 though so you don't
copy when clicking rather than dragging. The current check copies
text when it has a visible width of 1 but is actually multiple
characters in the rope like a CRLF line-ending. With this change
we check the unicode width of the character(s) in the selection
rather than the range length, so clicking on a CRLF line-ending
does not copy.
* use range.fragment to simplify getting the primary selection width
* redetect indent and line endings after language server replaces document
* removes nested if
* always redetect indent and line endings after format
This reverts commit 764d14f55894dc7213e48022dfa0f91829b8ef59.
The scrollbar length used to increase with more entries in the menu,
which was counter-intuitive to how scrollbars worked in most
applications. Turns out there was a typo in the floor division
implementation :)
The command palette previously used + as a delimiter for denoting
a single key in a key sequence, (like C+w). This was at odds with
how the statusline displayed them with pending keys (like <C-w>).
This patch changes the palette formatting to the statusline formatting
* feat: make `move_vertically` aware of tabs and wide characters
* refactor: replace unnecessary checked_sub with comparison
* refactor: leave pos_at_coords unchanged and introduce separate pos_at_visual_coords
* style: include comment to explain `pos_at_visual_coords` breaking condition
* refactor: use `pos_at_visual_coords` in `text_pos_at_screen_coords`
* feat: make `copy_selection_on_line` aware of wide characters
* Default rulers color to red
Currently if the theme a user is using doesn't have `ui.virtual.rulers`
set and they set up a ruler it just fails silently making it really hard
to figure out what went wrong. Did they set incorrectly set the ruler?
Are they using an outdated version of Helix that doesn't support rulers?
This happened to me today, I even switched to the default theme with
the assumption that maybe my theme just doesn't have the rulers setup
properly and it still didn't work.
Not sure if this is a good idea or not, feel free to suggest better
alternatives!
* Use builtin Style methods instead of Bevy style defaults
Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
* Only default the style if there's no ui or ui.virtual
* Update themes style from ui.virtual to ui.virtual.whitespace
* Revert ui.virtual change in onelight theme
* Prefer unwrap_or_else
Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
During write-quit, if the file fails to be written for any reason, helix
will still quit without saving the changes. This fixes this behavior by
introducing fallibility to the asynchronous job queues. This will also
benefit all contexts which may depend on these job queues.
Fixes#1575
When a new View of a Document is created, a default cursor of 0, 0 is
created, and it does not get normalized to a single width cursor until
at least one movement of the cursor happens. This appears to have no
practical negative effect that I could find, but it makes tests difficult
to work with, since the initial selection is not what you expect it to be.
This changes the initial selection of a new View to be the width of the
first grapheme in the text.
* Use new macro syntax for encoding sequences of keys
* Make convenience helpers for common test pattern
* Use indoc for inline indented raw strings
* Add feature flag for integration testing to disable rendering
- Add file-picker.follow-symlinks configuration option (default is true), this
also controls if filename and directory completers follow symlinks.
- Update FilePicker to set editor error if opening a file fails, instead of
panicing.
Fix#1548Fix#2246
In certain circumstances it was possible to get into an infinite loop
when replaying macros such as when different macros attempt to replay
each other.
This commit adds changes to track which macros are currently being
replayed and prevent getting into infinite loops.
When a goto command is cancelled, the jumplist should remain unchanged.
This commit delays saving the current selection to the jumplist until
jumping to a reference.
* Add shrink equivalent of extend_to_line_bounds
* Add a check for being past rope end in end position calc
* Include the EOL character in calculations
* Bind to `A-x` for now
* Document new keybind
* add Tree::swap_split_in_direction()
* add swap_view_{left,down,up,right} commands, bound to H,J,K,L
respectively in the Window menu(s)
* add test for view swapping
* feat(theme): add separate diagnostic colors
This commit adds separate diagnostic highlight colors for the different
types of LSP severities. If the severity type doesn't exist or is
unknown, we use some fallback coloring which was in use before this
commit.
Some initial color options were also added in the theme.toml
Resolves issue #2157
* feat(theme): add docs for new diagnostic options
* feat(theme): adjust defaults & reduce redundancy
- the different colors for different diagnostic severities are now
disabled in the default theme, instead diagnostics are just generally
underlined (as prior to the changes of this feature)
- the theme querying is now done once instead of every iteration in the
loop of processing every diagnostic message
* support insert register in prompt
* use next_char_handler instead of a flag
* Fix clippy issue
* show autoinfo when inserting register
* Revert "show autoinfo when inserting register"
This reverts commit 5488344de1.
* use completion instead of autoinfo
autoinfo is overlapped when using prompt
* recalculate_completion after inserting register
* Update helix-term/src/ui/prompt.rs
Co-authored-by: Ivan Tham <pickfire@riseup.net>
Co-authored-by: Ivan Tham <pickfire@riseup.net>
Change the layout of existing split view from horizontal to vertical and
vica-versa. It only effects the focused view and its siblings, i.e. not
recursive.
Command is mapped to 't' or 'C-t' under the Window menus.
This made sense initially when the implementation was still new (so we
got user reports more frequently), but a parsing error now generally
signifies a language server isn't properly implementing the spec.
Currently ctrl-w in insert mode deletes the cursor which results in
unexpected behavior. The patch also reduces the selection to cursor before
performing prev word to remove the behavior of removing unnecessary text
when nothing should be removed.
1. `::#(|)#::` after `ctrl-w` should be `#(|)#::`, previously `#(|)#:`
2. `#(|::)#` after `ctrl-w` should be `#(|::)#`, previously `#(|)#`
Fix#2390
Inserting a newline currently collapses any connected selections when inserting
or appending. It's happening because we're reducing the selections down to
their cursors (`let selection = ..` line) and then computing the new selection
based on the cursor. We're discarding the original head and anchor information
which are necessary to emulate Kakoune's behavior.
In Kakoune, inserting a newline retains the existing selection and _slides_
it (moves head and anchor by the same amount) forward by the newline and
indentation amount. Appending a newline extends the selection to include the
newline and any new indentation.
With the implementation of insert_newline here, we slide by adding the global
and local offsets to both head and anchor. We extend by adding the global
offset to both head and anchor but the local offset only to the head.
* Making the 'set-option' command help more descriptive.
* Adding the generated docs
* Making the message multi-line
* Replace newline with break in generated docs
* add reflow command
Users need to be able to hard-wrap text for many applications, including
comments in code, git commit messages, plaintext documentation, etc. It
often falls to the user to manually insert line breaks where appropriate
in order to hard-wrap text.
This commit introduces the "reflow" command (both in the TUI and core
library) to automatically hard-wrap selected text to a given number of
characters (defined by Unicode "extended grapheme clusters"). It handles
lines with a repeated prefix, such as comments ("//") and indentation.
* reflow: consider newlines to be word separators
* replace custom reflow impl with textwrap crate
* Sync reflow command docs with book
* reflow: add default max_line_len language setting
Co-authored-by: Vince Mutolo <vince@mutolo.org>
Allow tab-completion to continue when there is only a single, unambigous
completion target which is a directory. This allows e.g. nested directories
to be quickly drilled down just by hitting <tab> instead of first selecting
the completion then hitting <enter>.
* add run_shell_command
* docgen
* fix command name
Co-authored-by: Blaž Hrastnik <blaz@mxxn.io>
* refactored Info::new
* show 'Command failed' if execution fails
* TypedCommand takes care of error handling and printing the error to the statusline.
* docgen
* use Popup instead of autoinfo
* remove to_string in format!
* Revert chage in info.rs
* Show "Command succeed" when success
* Fix info.rs
Co-authored-by: Blaž Hrastnik <blaz@mxxn.io>
When fiddling with paths in a :o prompt, one usually would want Ctrl-W to erase a path segment
rather than the whole path. This is how Ctrl-W works in e.g. (neo)vim out of the box.
Currently A-left move one word left and the behavior will be more
consistent for people coming GUI world if the key was changed to control
given that both browsers and editors like vscode uses C-left right by
default to move word rather than alt.
A-hl currently is not very consistent with hl when next object is
selected, since it may go up/down or left/right and this behavior is
confusing such that some people think it should swap the keys with A-jk,
so it is better to use A-pn since that only specifies two direction.
A-jk have the same issue as in it usually moves right and is not
consistent with the behavior of jk so people may think A-hl is better,
maybe A-oi is better here since A-hl will be swapped to A-pn, A-oi can
convey the meaning of in and out, similar to some window manager keys?
* feat(commands): better handling of buffer-close
Previously, when closing buffer, you would loose cursor position in other docs.
Also, all splits where the buffer was open would be closed.
This PR changes the behavior, if the view has also other buffer
previously viewed it switches back to the last one instead of the view
being closed. As a side effect, since the views are persisted,
the cursor history is persisted as well.
Fixes: https://github.com/helix-editor/helix/issues/1186
* Adjust buffer close behavior
* Remove closed documents from jump history
* Fix after rebase
* Fix ctrl-u on insert behavior
Now should follow vim behavior more
- no longer remove text on cursor
- no longer remove selected text while inserting
- first kill to start non-whitespace, start, previous new line
* Add comment for c-u parts
Select multiple line and open should be based on the whole selection
and not just the line of the cursor, which causes weird behavior like
opening in the middle of the selection which user might not expect.
* added command to extend selection to line above
* fixed view not scrolling up when reaching top of the screen
* refactored shared code into separate impl
* Send active diagnostics to LSP when requesting code actions.
This allows for e.g. clangd to properly send the quickfix code actions
corresponding to those diagnostics as options.
The LSP spec v3.16.0 introduced an opaque `data` member that would allow
the server to persist arbitrary data between the diagnostic and the code
actions request, but this is not supported yet by this commit.
* Reuse existing range_to_lsp_range functionality
hx --health output table's second and third columns were not showing
symbols like ✔ or ✘ to indicate whether LSP or DAP binaries were found.
This change adds these symbols to improve accessibility.
Fixes#1894
Signed-off-by: Nirmal Patel <npate012@gmail.com>
* Make `:write` create nonexistent subdirectories
Prompting as to whether this should take place remains a TODO.
* Move subdirectory creation to new `w!` command
The `file_picker_at_current_directory` command opens the file picker at
the current working directory (CWD). This can be useful when paired with
the built-in `:cd` command which changes the CWD.
It has been mapped to `space F` by default.
During a single HighlightEvent::Source, the highlight spans do not
change and we can merge them into a single style at the beginning
of the event and use it instead of re-computing it for every grapheme
* Add runtime language configuration (#1794)
* Add set-language typable command to change the language of current buffer.
* Add completer for available language options.
* Update set-language to refresh language server as well
* Add language id based config lookup on `syntax::Loader`.
* Add `Document::set_language3` to set programming language based on language
id.
* Update `Editor::refresh_language_server` to try language detection only if
language is not already set.
* Remove language detection from Editor::refresh_language_server
* Move document language detection to where the scratch buffer is saved.
* Rename Document::set_language3 to Document::set_language_by_language_id.
* Remove unnecessary clone in completers::language
when using helix over mosh, the screen doesn't get cleared and
characters get left all over the place until they are overwritten. with
this change, the screen gets properly cleared as soon as helix starts
The search implementation would start searching at the next grapheme
boundary after the previous selection. In case the next occurence of the
needle is immediately after the current selection, this occurence would
not be found (without wraparound) because the first grapheme is skipped.
The correct approach is to use the ensure_grapheme_boundary functions instead
of using the functions that skip unconditionally to the next grapheme.
Currently, the picker's re-using a few bindings which are also present
in the prompt. This causes some editing behaviours to not function on
the picker.
**Ctrl + k** and **Ctrl + j**
This should kill till the end of the line on prompt, but is overridden
by the picker for scrolling. Since there are redundancies (`Ctrl + p`,
`Ctrl + n`), we can remove it from picker.
**Ctrl + f** and **Ctrl + b**
This are used by the prompt for back/forward movement. We could modify
it to be Ctrl + d and Ctrl + u, to match the `vim` behaviour.
* WIP: Rework indentation system
* Add ComplexNode for context-aware indentation (including a proof of concept for assignment statements in rust)
* Add switch statements to Go indents.toml (fixes the second half of issue #1523)
Remove commented-out code
* Migrate all existing indentation queries.
Add more options to ComplexNode and use them to improve C/C++ indentation.
* Add comments & replace Option<Vec<_>> with Vec<_>
* Add more detailed documentation for tree-sitter indentation
* Improve code style in indent.rs
* Use tree-sitter queries for indentation instead of TOML config.
Migrate existing indent queries.
* Add documentation for the new indent queries.
Change xtask docgen to look for indents.scm instead of indents.toml
* Improve code style in indent.rs.
Fix an issue with the rust indent query.
* Move indentation test sources to separate files.
Add `#not-kind-eq?`, `#same-line?` and `#not-same-line` custom predicates.
Improve the rust and c indent queries.
* Fix indent test.
Improve rust indent queries.
* Move indentation tests to integration test folder.
* Improve code style in indent.rs.
Reuse tree-sitter cursors for indentation queries.
* Migrate HCL indent query
* Replace custom loading in indent tests with a designated languages.toml
* Update indent query file name for --health command.
* Fix single-space formatting in indent queries.
* Add explanation for unwrapping.
Co-authored-by: Triton171 <triton0171@gmail.com>
* publish a source tarball with version and grammars
* include_str! the release version from a VERSION file
* remove setting of .version file from tag
don't need this anymore since the file is checked into source
* Move top level lsp config to editor.lsp
This is mainly done to accomodate the new lsp.signature-help config
option that will be introduced in https://github.com/helix-editor/helix/pull/1755
which will have to be accessed by commands. The top level config
struct is split and moved to different places, making the relocation
necessary
* Revert rebase slipup
This avoids costly conversions via byte_to_char (which are then
reversed back into bytes internally in Ropey).
Reduces time spent in slice/byte_to_char from ~24% to ~5%.
When the picker results output is empty, movement actions result in a panic:
```
thread 'main' panicked at 'attempt to calculate the remainder with a divisor of zero', helix-term/src/ui/picker.rs:420:31
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
```
This could be a no-op instead when the matches length is zero.
Currently match is finding the match based on the anchor rather than the
head (cursor) so this behavior is rather unexpected when user is doing
a match but a different item was matched instead when the selection is
more than one character.
This restores much of the behavior that existed before this PR:
helix will build the grammars when compiling. The difference is that
now fetching is also done during the build phase and is done much
more quickly - both shallow and in parallel.
This is a rather large refactor that moves most of the code for
loading, fetching, and building grammars into a new helix-loader
module. This works well with the [[grammars]] syntax for
languages.toml defined earlier: we only have to depend on the types
for GrammarConfiguration in helix-loader and can leave all the
[[language]] entries for helix-core.
The vision with 'use-grammars' is to allow the long-requested feature
of being able to declare your own set of grammars that you would like.
A simple schema with only/except grammar names controls the list
of grammars that is fetched and built. It does not (yet) control which
grammars may be loaded at runtime if they already exist.
build_grammars adapts the functionality that previously came from
helix-syntax to be used at runtime from the command line flags.
fetch_grammars wraps command-line git to perform the same actions
previously done in the scripts in #1560.
helix-syntax mostly existed for the sake of the build task which
checks and compiles the submodules. Since we won't be relying on
that process anymore, it doesn't end up making much sense to have
a very thin crate just for some functions that we could port to
helix-core.
The remaining build-related code is moved to helix-term which will
be able to provide grammar builds through the --build-grammars CLI
flag.
* Move runtime file location definitions to core
* Add basic --health command
* Add language specific --health
* Show summary for all langs with bare --health
* Use TsFeature from xtask for --health
* cargo fmt
Co-authored-by: Blaž Hrastnik <blaz@mxxn.io>
* Add arrow key mappings for tree-sitter parent/child/sibling nav
This helps my use case, where I use a non-qwerty layout with a
programmable mechanical keyboard, and use a layer switching key (think
fn) to send left down up right from the traditional hjkl positions.
* Add new bindings to docs
* Fix bug with auto replacing components in compositor
This was last known to be working with 5995568c at the
time of commit, but now doesn't work with latest rust
stable.
The issue probably stems from using
std::any::type_name() for finding a component in the
compositor, for which the docs explicitly warn against
considering it as a unique identifier for types.
`replace_or_push()` takes a boxed `Component` and
passes it to `find_id()` which compares this with a
bare Component. `type_name()` returns `Box<T>` for
the former and `T` for latter and we have a false
negative. This has been solved by using a generics
instead of trait objects to pass in a `T: Component`
and then use it for comparison.
I'm not exactly sure how this worked fine at the
time of commit of 5995568c; maybe the internal
implementation of `type_name()` changed to properly
indicate indirection with Box.
* Do not compare by type name in compositor find_id
* add basic completion replay
* use transaction as the last completion
* completion replay only on trigger position
* cache changes in CompletionAction
Co-authored-by: Blaž Hrastnik <blaz@mxxn.io>
* Implement buffer-close-all
* Implement buffer-close-others
* Refactor all buffer close variants to use shared logic
* Fix clippy lint
* Docgen for new commands
* Shorten error message for attempting to close buffers that don't exist
* Refactor shared buffer methods to pass only editor, not whole compositor
* Switch signature of bulk buffer closing to use slice of DocumentIds
Addresses feedback that accepting an IntoIterator implementor is too
much for an internal. Also possibly saves some moving?
* Show infobox to hint textobjects with `mi` and `ma`
* Add note to infobox than any pair of characters will work too
The wording could probably be a little more clear, but I wanted to
keep it short but still accurate.
* Don't allocate a vec for the static help text
* Fix bug where `mi<esc>` would swallow next input and persist infobox
* Better help text for arbitrary pair matching in textobject selection
* Add way to add fake pending key data below status, use with `mi`/`ma`
This is a bit hacky as it makes use of global state which will end
up managed in multiple places, but has precedent in the way autoinfo
works. There should probably be a bigger refactor to handle this
kind of state better.
* Return early on anything other than `mi` and `ma` for autoinfo
* Remove "ascii" from help text with `mi` and `ma`
* Update helix-term/src/ui/editor.rs
Co-authored-by: Blaž Hrastnik <blaz@mxxn.io>
* Refactor file picker filetype filter logic to remove panic, make clearer
An unwrap was unneccesarily present due to a prior contribution of mine
which was before I had any understanding of error handling in Rust. I've
also swapped a match for an if let, as was originally suggested in the
original pull request adding filetype filtering, but was merged before I
could address.
* Add some comments to the file picker code for clarity
* Switch to expect instead of ignoring type def error
* ignore Enter keypress when menu has no selection
supersedes #1622
Builds on the work in #1285. I want to allow Enter to create a newline
when there is no selection in the autocomplete menu.
This occurs somewhat often when using LSP autocomplete in Elixir which
uses `do/end` blocks (and I set the autocomplete menu delay to 0 which
exacerbates the problem):
```elixir
defmodule MyModule do
def do_foo(x) do
x
end
def other_function(y) do|
end
```
Here the cursor is `|` in insert mode. The LSP suggests `do_foo` but I
want to create a newline. Hitting Enter currently closes the menu,
so I end up having to hit Enter twice when the module contains any
local with a `do` prefix, which can be inconsistent. With this change,
we ignore the Enter keypress to end up creating the newline in this case.
* pop compositor layer when ignoring Enter keypress
* move closing function out of consumed event result closure
* explicitly label close_fn as an 'Option<Callback>'
* impl auto pairs config
Implements configuration for which pairs of tokens get auto completed.
In order to help with this, the logic for when *not* to auto complete
has been generalized from a specific hardcoded list of characters to
simply testing if the next/prev char is alphanumeric.
It is possible to configure a global list of pairs as well as at the
language level. The language config will take precedence over the
global config.
* rename AutoPair -> Pair
* clean up insert_char command
* remove Rc
* remove some explicit cloning with another impl
* fix lint
* review comments
* global auto-pairs = false takes precedence over language settings
* make clippy happy
* print out editor config on startup
* move auto pairs accessor into Document
* rearrange auto pair doc comment
* use pattern in Froms
* Add Event::Used to use event callback without consuming
* Close popup if contents ignored event
* collect event results before executing callbacks
* don't add new result variant, use Ignored(..) instead
* break in match cases
* Make auto_close configurable
* fix merge
* auto close hover popups
* fix formatting
Some users (including myself) want to turn off filtering of files
prefixed with `.`, as they are often useful to edit. For example, `.env`
files, configuration for linters `.eslint.json` and the like.
The unwrap (or '.ok()' rather) triggers for some errors but not
negative status codes. In the case where helix is being packaged
in an empty git repository, the existing mechanism will fail because
git init
git rev-parse HEAD
gives a negative exit code and prints to stderr
stderr: "fatal: ambiguous argument 'HEAD': unknown revision or path not in the working tree....
with a stdout of "HEAD\n" (too short to slice with [..8]).
* feat(commands): command palette
Add new command to display command pallete that can be used
to discover and execute available commands.
Fixes: https://github.com/helix-editor/helix/issues/559
* Make picker take the whole context, not just editor
* Bind command pallete
* Typable commands also in the palette
* Show key bindings for commands
* Fix tests, small refactor
* Refactor keymap mapping, fix typo
* Ignore sequence key bindings for now
* Apply suggestions
* Fix lint issues in tests
* Fix after rebase
Co-authored-by: Blaž Hrastnik <blaz@mxxn.io>
In order to implement this completer, the completion function needs to
be able to access the compositor's context (to allow it to get the
list of buffers currently open in the context's editor).
If building from source and the source is contained in a larger
repository, we'd contain the wrong version. It's also easy to
accidentally have a newer tag that would change the version.
This code:
let start = ensure_grapheme_boundary_next(text, text.byte_to_char(start));
let end = ensure_grapheme_boundary_next(text, text.byte_to_char(end));
Would convert byte to char index, but then internally immediately convert back
to byte index, operate on it, then convert it to char index.
This change reduces the amount of time spent in ensure_grapheme_boundary from
29% to 2%.
* add select_next_sibling and select_prev_sibling commands
* refactor objects to use higher order functions
* address clippy feedback
* move selection cloning into commands
* add default keybindings under left/right brackets
* use [+t,]+t for selecting sibling syntax nodes
* setup Alt-{j,k,h,l} default keymaps for syntax selection commands
* reduce boilerplate of select_next/prev_sibling in commands
* import tree-sitter Node type in commands
* add show_subtree command for viewing tree-sitter subtree in Popup
* remove '.slice(..)' from show_subtree command
* name docs and subtree Popups 'hover'
* feat(commands): shrink_selection
Add `shrink_selection` command that can be used to shrink
previously expanded selection.
To make `shrink_selection` work it was necessary to add
selection history to the Document since we want to shrink
the selection towards the syntax tree node that was initially
selected.
Selection history is cleared any time the user changes
selection other way than by `expand_selection`. This ensures
that we don't get some funky edge cases when user calls
`shrink_selection`.
Related: https://github.com/helix-editor/helix/discussions/1328
* Refactor shrink_selection, move history to view
* Remove useless comment
* Add default key mapping for extend&shrink selection
* Rework contains_selection method
* Shrink selection without expand selects first child
* feat(commands): ensure_selections_forward
Add command that ensures that selections are in forward direction.
Fixes: https://github.com/helix-editor/helix/issues/1332
* Add keybinding for ensure_selections_forward
Add `A-:` keybinding for the ensure_selections_forward command.
* Re-use range.flip for flip_selections command
* feat: Update settings at runtime
fix the clippy warning
* update the documentation
* use to_value instead of to_vec+from_value
* drop the equal
* remove an useless comment
* apply suggestion
* feat(ui): file encoding in statusline
Display file encoding in statusline if the encoding
isn't UTF-8.
* Re-export encoding_rs from core
From there it can be imported by other mods
that rely on it.
* feat(lsp): configurable diagnostic severity
Allow severity of diagnostic messages to be configured.
E.g. allow turning of Hint level diagnostics.
Fixes: https://github.com/helix-editor/helix/issues/1007
* Use language_config() method
* Add documentation for diagnostic_severity
* Use unreachable for unknown severity level
* fix: documentation for diagnostic_severity config
* tmp add code for dedent
* finish normal_mode with dedent behavior
* use function pointer
* rebase from origin
* check dedent condition inside normal_mode implementation
* using if let...
* fix check
* using char_is_whitespace instead of ch.is_whitespace
* fix clippy
* abstract restore_indent function
* Add auto pairs for same-char pairs
* Add unit tests for all existing functionality
* Add auto pairs for same-char pairs (quotes, etc). Account for
apostrophe in prose by requiring both sides of the cursor to be
non-pair chars or whitespace. This also incidentally will work for
avoiding a double single quote in lifetime annotations, at least until
<> is added
* Slight factor of moving the cursor transform of the selection to
inside the hooks. This will enable doing auto pairing with selections,
and fixing the bug where auto pairs destroy the selection.
Fixes#1014
* Move `runtime/themes/base16_default_terminal.toml` to
`base16_theme.toml` alongside `theme.toml`
* Use `terminfo` crate to detect whether the terminal supports true
color and, if the user has no theme configured and their terminal does
not support true color, load the alt default theme instead of the
normal default.
Remove `terminfo` dependency, use `COLORTERM` env instead
Prevent user from switching to an unsupported theme
Add `true-color-override` option
If the terminal is wrongly detected to not support true color,
`true-color-override = true` will override the detection.
Rename `true-color-override` to `true-color`
* Macros WIP
`helix_term::compositor::Callback` changed to take a `&mut Context` as
a parameter for use by `play_macro`
* Default to `@` register for macros
* Import `KeyEvent`
* Special-case shift-tab -> backtab in `KeyEvent` conversion
* Move key recording to the compositor
* Add comment
* Add persistent display of macro recording status
When macro recording is active, the pending keys display will be shifted
3 characters left, and the register being recorded to will be displayed
between brackets — e.g., `[@]` — right of the pending keys display.
* Fix/add documentation
Following the addition of `xtask`, `cargo run` has multiple possible
targets, necessitating the usage of `cargo run --bin hx` to run Helix
during development. This allows `cargo run` to be used to run `hx`.
1. pressing o on a line with no indentation will open a new line as
expected, but esc will then delete the line altogether
2. the kill_line behavior happens after insert mode changes are already
commited to history, and the change isn't commited. pressing u after
this will break highlighting & undo history
This reverts commit c08d2fae58.
Tree sitter returns an index referring to the position of the scope in
the scopes array. We can use that same index to avoid a hashmap lookup
and instead store the styles in an array.
This currently stores the styles in both a map and an array because the
UI still uses hashmap lookups, but it's a reasonable tradeoff.
* restore indent when press esc right after open a new line
* add comment for restore_indent
* fix, and make kill to line end behaves like emacs
* update comment
* fix comment
* adjust cancel restore_indent situation
* check esc logic in mode transaction
* improve comment
* add more check for dedent
* update comment
* use matches to check for last_cmd
* no need to introduct CommandFun type
* accept count for goto_window
also fix view is not fullfilled issue
* fix fulfilled mispell
* Update helix-term/src/commands.rs
Co-authored-by: Ivan Tham <pickfire@riseup.net>
* Update helix-term/src/commands.rs
Co-authored-by: Ivan Tham <pickfire@riseup.net>
* fix merge issue
* revert line computation logic
Co-authored-by: Ivan Tham <pickfire@riseup.net>
* goto_file
* support goto_file under current cursor
* add C-w f/F
* sync space w with window mode
* Update helix-term/src/commands.rs
Co-authored-by: Blaž Hrastnik <blaz@mxxn.io>
fixes#1136
* removed a log::info
* removed temp.rs
* cargo clippy no longer complains
* new get_lang_server function
* get_lang_server is now launch_language_server
* launch_lang_server will now close the previous one
* better code readability
* remove resfresh_ls(and a wrong comment)
* Do not crash when run goto command without line number
Report an error when running goto command without entering a
line number.
Fixes#1159
* Use is_empty() instead check len zero
* Add typable `goto` command
* Support `:<line-number>` on prompt
* Rename function according to convention
* Directly call into goto_line_number function
* align lines
* remove log statement
* use selections to align
* fix a clippy issue
* only accept 1,2,3 as user count
* Update helix-term/src/commands.rs
Co-authored-by: Ivan Tham <pickfire@riseup.net>
* return if user count is not correct
* add doc
Co-authored-by: Ivan Tham <pickfire@riseup.net>
* squashed WIP commits
* hide_gitignore working with config
* pass reference to new config parameter of file_picker()
* update config option name to match name on walk builder
* add comments to config and documentation of option to book
* add git_ignore option to WalkBuilder within prompt in commands.rs
* WIP: add FilePickerConfig struct
* WIP: cleanup
* WIP: add more options including max_depth
* WIP: changed defaults to match ignore crate defaults
* WIP: change WalkBuilder in global_search() to use config options
* WIP: removed follow_links, changed max_depth to follow config setting
* WIP: update book with file-picker inline table notation
* update documentation for file-picker config in book
* adjusted to [editor.file-picker] in book configuration.md
* adjust comments in editor.rs to be doc comments, cleanup
* adjust comments
* adjust book
* Jump to end char of surrounding pair from any cursor pos
* Separate bracket matching into exact and fuzzy search
* Add constants for bracket chars
* Abort early if char under cursor is not a bracket
* Simplify bracket char validation
* Refactor node search and unify find methods
* Remove bracket constants
