Currently in Zed, certain characters require pressing the key twice to
move the caret through that character. For example: "❤️" and "y̆".
The reason for this is as follows:
Currently, Zed uses `chars` to distinguish different characters, and
calling `chars` on `y̆` will yield two `char` values: `y` and `\u{306}`,
and calling `chars` on `❤️` will yield two `char` values: `❤` and
`\u{fe0f}`.
Therefore, consider the following scenario (where ^ represents the
caret):
- what we see: ❤️ ^
- the actual buffer: ❤ \u{fe0f} ^
After pressing the left arrow key once:
- what we see: ❤️ ^
- the actual buffer: ❤ ^ \u{fe0f}
After pressing the left arrow key again:
- what we see: ^ ❤️
- the actual buffer: ^ ❤ \u{fe0f}
Thus, two left arrow key presses are needed to move the caret, and this
PR fixes this bug (or this is actually a feature?).
I have tried to keep the scope of code modifications as minimal as
possible. In this PR, Zed handles such characters as follows:
- what we see: ❤️ ^
- the actual buffer: ❤ \u{fe0f} ^
After pressing the left arrow key once:
- what we see: ^ ❤️
- the actual buffer: ^ ❤ \u{fe0f}
Or after pressing the delete key:
- what we see: ^
- the actual buffer: ^
Please note that currently, different platforms and software handle
these special characters differently, and even the same software may
handle these characters differently in different situations. For
example, in my testing on Chrome on macOS, GitHub treats `y̆` as a
single character, just like in this PR; however, in Rust Playground,
`y̆` is treated as two characters, and pressing the delete key does not
delete the entire `y̆` character, but instead deletes `\u{306}` to yield
the character `y`. And they both treat `❤️` as a single character,
pressing the delete key will delete the entire `❤️` character.
This PR is based on the principle of making changes with the smallest
impact on the code, and I think that deleting the entire character with
the delete key is more intuitive.
Release Notes:
- Fix caret movement issue for some special characters
---------
Co-authored-by: Conrad Irwin <conrad.irwin@gmail.com>
Co-authored-by: Thorsten <thorsten@zed.dev>
Co-authored-by: Bennet <bennetbo@gmx.de>
This PR fixes a potential panic that could occur when loading malformed
Wasm files.
We now use the `parse_wasm_extension_version` function that was
previously used just to extract the Zed extension API version from the
Wasm bytes as a pre-validation step. By parsing the entirety of the Wasm
file here instead of returning as soon as we find the version, the
invalid Wasm bytes are now surfaced as an `Err` instead of a panic.
We were able to replicate the panic using the following test:
```rs
#[gpui::test]
async fn test_bad_wasm(cx: &mut TestAppContext) {
init_test(cx);
let wasm_host = cx.update(|cx| {
WasmHost::new(
FakeFs::new(cx.background_executor().clone()),
FakeHttpClient::with_200_response(),
FakeNodeRuntime::new(),
Arc::new(LanguageRegistry::test(cx.background_executor().clone())),
PathBuf::from("/the/work/dir".to_string()),
cx,
)
});
let mut wasm_bytes = std::fs::read("/Users/maxdeviant/Library/Application Support/Zed/extensions/installed/dart/extension.wasm").unwrap();
// This is the error message we were seeing in the stack trace:
// range end index 267037 out of range for slice of length 253952
dbg!(&wasm_bytes.len());
// Truncate the bytes to the same point:
wasm_bytes.truncate(253952);
std::fs::write("/tmp/bad-extension.wasm", wasm_bytes.clone()).unwrap();
let manifest = Arc::new(ExtensionManifest {
id: "the-extension".into(),
name: "The Extension".into(),
version: "0.0.1".into(),
schema_version: SchemaVersion(1),
description: Default::default(),
repository: Default::default(),
authors: Default::default(),
lib: LibManifestEntry {
kind: None,
version: None,
},
themes: Default::default(),
languages: Default::default(),
grammars: Default::default(),
language_servers: Default::default(),
});
// 💥
let result = wasm_host
.load_extension(wasm_bytes, manifest, cx.executor())
.await;
dbg!(result.map(|_| ()));
```
Release Notes:
- Fixed a crash that could occur when loading malformed Wasm extensions
([#10352](https://github.com/zed-industries/zed/issues/10352)).
---------
Co-authored-by: Max <max@zed.dev>
There hasn't been a componentized way to create inputs or text fields
thus far due to the innate circular dependency between the `ui` and
`editor` crates. To bypass this issue we are introducing a new
`ui_text_field` crate to specifically handle this component.
`TextField` provides the ability to add stacked or inline labels, as
well as applies a standard visual style to inputs.
Example:
We'll continue to evolve this component in the near future and start
using it in the app once we've built out the needed functionality.
Release Notes:
- N/A
---------
Co-authored-by: Marshall Bowers <elliott.codes@gmail.com>
This panic has occured a handful of times, I think it must be the case
that:
1. Item is dropped outside of an update loop
2. The next update is this one
In that case no flush effects will have called the release observers
yet, but we cannot upgrade the WeakModel because the ref count is 0
Release Notes:
- Fixed a (rare) panic while collaborating
This PR renames `language::Buffer::new` to `language::Buffer::local` and
simplifies its interface. Instead of taking a replica id (which should
always be 0 for the local case) and a `BufferId`, which was awkward and
verbose to construct, it simply takes text and a `cx`.
It uses the `cx` to derive a `BufferId` from the `EntityId` associated
with the `cx`, which should always be positive based on the following
analysis...
We convert the entity id to a u64 using this method on `EntityId`, which
is defined by macros in the `slotmap` crate:
```rust
pub fn as_ffi(self) -> u64 {
(u64::from(self.version.get()) << 32) | u64::from(self.idx)
}
```
If you look at the type of `version` in `KeyData`, it is non-zero:
```rust
#[derive(Clone, Copy, PartialEq, Eq, PartialOrd, Ord, Hash)]
pub struct KeyData {
idx: u32,
version: NonZeroU32,
}
```
This commit also adds `Context::reserve_model` and
`Context::insert_model` to determine a model's entity ID before it is
created, which we need in order to assign a `BufferId` in the background
when loading a buffer asynchronously.
Release Notes:
- N/A
---------
Co-authored-by: Piotr Osiewicz <24362066+osiewicz@users.noreply.github.com>
- Fixed#8603
For the label title of the project panel, I find that there is no place
to use to get the title of the label to do some operations, it should be
safe to modify it, but I'm not sure how we need to modify the problem, I
can think of two scenarios:
1. Modify every place where you don't want multiple lines to appear
2. Make the label only display a single line (e.g. provide a new
parameter, or a new label type?)
Followup to #10327
It can be enabled with the following setting:
"line_indicator_format": "short"
No release note, as the original change didn't go out to Preview yet.
/cc @bartekpacia @0x2CA
Release Notes:
- N/A
This fixes certain shortcuts/motions on X11 like in vim mode "v i )",
where previously zed would interpret it as "v i SHIFT )" due to the x11
backend emitting key press events for modifier keys even though other
platforms like Wayland don't. This also adds support for
ModifiersChanged events to X11
Release Notes:
- Fixed vim motions like "v i )" not working on X11
([#10199](https://github.com/zed-industries/zed/issues/10199)).
Release Notes:
- N/A
Picks up https://github.com/kvark/blade/pull/105,
https://github.com/kvark/blade/pull/97, and more
Switches the presentation to be non-blocking, which will improve the
latency slightly.
Allows to start playing with GLES backend, e.g.
```bash
cd crates/gpui
RUSTFLAGS="--cfg gles" CARGO_TARGET_DIR=./target-gl cargo run --example hello_world
```
It doesn't currently render properly due to an issue that needs
investigation, see
https://github.com/kvark/blade/pull/105#issuecomment-2041006542
But at least it's a start
Co-authored-by: Mikayla Maki <mikayla@zed.dev>
This fixes a bug in https://github.com/zed-industries/zed/pull/9818,
where the status was not removed if the request failed. It also adds
replication of these new status messages to guests when collaborating.
Release Notes:
- Fixed an issue where the status of failed LSP actions was left in the
status bar
---------
Co-authored-by: Marshall <marshall@zed.dev>
Co-authored-by: Marshall Bowers <elliott.codes@gmail.com>
This fixed an issue introduced in
https://github.com/zed-industries/zed/pull/10126, where, when toggling
comments in a language with multiple line comment prefixes (e.g. Gleam,
Erlang) Zed would insert the *last* prefix instead of the first.
Release Notes:
- Fixed an issue where the `toggle comments` command inserted the wrong
line comment prefix in some languages (preview only).
Co-authored-by: Marshall <marshall@zed.dev>
Part of https://github.com/zed-industries/zed/issues/10132 extracted out
of a bigger PR that refactors tasks and shows labels that look more like
commands (ergo needs better readable package names)
Release Notes:
- N/A
This should help with some of the memory problems reported in
https://github.com/zed-industries/zed/issues/8436, especially the ones
related to large files (see:
https://github.com/zed-industries/zed/issues/8436#issuecomment2037442695),
by **reducing the memory required to represent a buffer in Zed by
~50%.**
### How?
Zed's memory consumption is dominated by the in-memory representation of
buffer contents.
On the lowest level, the buffer is represented as a
[Rope](https://en.wikipedia.org/wiki/Rope_(data_structure)) and that's
where the most memory is used. The layers above — buffer, syntax map,
fold map, display map, ... — basically use "no memory" compared to the
Rope.
Zed's `Rope` data structure is itself implemented as [a `SumTree` of
`Chunks`](8205c52d2b/crates/rope/src/rope.rs (L35-L38)).
An important constant at play here is `CHUNK_BASE`:
`CHUNK_BASE` is the maximum length of a single text `Chunk` in the
`SumTree` underlying the `Rope`. In other words: It determines into how
many pieces a given buffer is split up.
By changing `CHUNK_BASE` we can adjust the level of granularity
withwhich we index a given piece of text. Theoretical maximum is the
length of the text, theoretical minimum is 1. Sweet spot is somewhere
inbetween, where memory use and performance of write & read access are
optimal.
We started with `16` as the `CHUNK_BASE`, but that wasn't the result of
extensive benchmarks, more the first reasonable number that came to
mind.
### What
This changes `CHUNK_BASE` from `16` to `64`. That reduces the memory
usage, trading it in for slight reduction in performance in certain
benchmarks.
### Benchmarks
I added a benchmark suite for `Rope` to determine whether we'd regress
in performance as `CHUNK_BASE` goes up. I went from `16` to `32` and
then to `64`. While `32` increased performance and reduced memory usage,
`64` had one slight drop in performance, increases in other benchmarks
and substantial memory savings.
| `CHUNK_BASE` from `16` to `32` | `CHUNK_BASE` from `16` to `64` |
|-------------------|--------------------|
|
|
|
### Real World Results
We tested this by loading a 138 MB `*.tex` file (parsed as plain text)
into Zed and measuring in `Instruments.app` the allocation.
#### standard allocator
Before, with `CHUNK_BASE: 16`, the memory usage was ~827MB after loading
the buffer.
| `CHUNK_BASE: 16` |
|---------------------|
|
|
After, with `CHUNK_BASE: 64`, the memory usage was ~396MB after loading
the buffer.
| `CHUNK_BASE: 64` |
|---------------------|
|
|
#### `mimalloc`
`MiMalloc` by default and that seems to be pretty aggressive when it
comes to growing memory. Whereas the std allocator would go up to
~800mb, MiMalloc would jump straight to 1024MB.
I also can't get `MiMalloc` to work properly with `Instruments.app` (it
always shows 15MB of memory usage) so I had to use these `Activity
Monitor` screenshots:
| `CHUNK_BASE: 16` |
|---------------------|
|
|
| `CHUNK_BASE: 64` |
|---------------------|
|
|
### Release Notes
Release Notes:
- Reduced memory usage for files by up to 50%.
---------
Co-authored-by: Antonio <antonio@zed.dev>
This PR updates the extensions dependent on v0.0.6 of the
`zed_extension_api` crate to use the now-published version on crates.io
instead of a path dependency.
The impacted extensions are:
- `dart`
- `gleam`
- `haskell`
- `svelte`
Release Notes:
- N/A
To reference the system font, use the special ".SystemUIFont" family
name.
/cc @PixelJanitor
Release Notes:
- Switched to the system UI font for user interface elements on macOS.
Co-authored-by: Antonio Scandurra <antonio@zed.dev>
This PR makes it so our temporary host-side workaround for setting
certain language server binaries as executable only applies to binaries
that are downloaded by the extension.
Previously we would do this for any binary, including ones that could
have been sourced from the $PATH.
Release Notes:
- Fixed a file permissions issue when trying to use a Zig language
server (`zls`) present on the $PATH.
Fixed#10149
A user had Zed crash due to invalid font size in settings. It turned out
the width/height of glyphs does not pass validation in Metal texture
initialization with a large enough font size.
All modern Macs have a max texture width/height of 16kB (barring Apple
A8, used by iPhone 6 back in 2014, which uses 8kB). This commit clamps
texture size at 16kB. Note that while it fixes Zed crash, using a font
size that hits the limit is still pretty unusable - the users will still
have a pretty unusable editor, but at least it won't crash for them.
Release Notes:
- Fixed crashes with huge `buffer_font_size` values.
This PR changes ways the Find/Replace functionality in the
Buffer/Project Search is accessible via shortcuts. It makes those panels
work the same way as in VS Code and Sublime Text.
The details are described in the issue: [Make Find/Replace easier to
use](https://github.com/zed-industries/zed/issues/9142)
There's a difficulty with the Linux keybindings:
VS Code uses on MacOS (this PR replicates it):
| Action | Buffer Search | Project Search |
| --- | --- | --- |
| Find | `cmd-f` | `cmd-shift-f` |
| Replace | `cmd-alt-f` | `cmd-shift-h` |
VS Code uses on Linux (this PR replicates all but one):
| Action | Buffer Search | Project Search |
| --- | --- | --- |
| Find | `ctrl-f` | `ctrl-shift-f` |
| Replace | `ctrl-h` ❗ | `ctrl-shift-h` |
The problem is that `ctrl-h` is already taken by the `editor::Backspace`
action in Zed on Linux.
There's two options here:
1. Change keybinding for `editor::Backspace` on Linux to something else,
and use `ctrl-h` for the "replace in buffer" action.
2. Use some other keybinding on Linux in Zed. This PR introduces
`ctrl-r` for this purpose, though I'm not sure it's the best choice.
What do you think?
fixes#9142
Release Notes:
- Improved access to "Find/Replace in Buffer" and "Find/Replace in
Files" via shortcuts (#9142).
Optionally, include screenshots / media showcasing your addition that
can be included in the release notes.
- N/A
This PR updates the Gleam extension to strip out newlines in the
completion details returned from the language server.
These newlines were causing the completion menu to compute a large
height for each item, resulting in lots of empty space in the completion
menu:
<img width="878" alt="Screenshot 2024-04-08 at 8 53 29 PM"
src="https://github.com/zed-industries/zed/assets/1486634/383c52ec-e5cb-4496-ae4c-28744b4ecaf5">
The approach to stripping newlines allocates a bit more than I would
like.
It would be good to see if it is possible for the Gleam language server
to not send us these newlines in the first place.
Release Notes:
- N/A
This PR flips the optionality of the `AutoUpdateSettingContent` to make
it a bit easier to work with.
#### Before
```rs
struct AutoUpdateSettingContent(Option<bool>);
type FileContent = AutoUpdateSettingContent;
```
#### After
```rs
struct AutoUpdateSettingContent(bool);
type FileContent = Option<AutoUpdateSettingContent>;
```
Release Notes:
- N/A
This puts the Linux platform implementation at a similar code style and
quality to the macOS platform. The largest change is that I collapsed
the `LinuxPlatform` -> `[Backend]` -> `[Backend]State` ->
`[Backend]StateInner` to just `[Backend]` and `[Backend]State`, and in
the process removed most of the `Rc`s and `RefCell`s.
TODO:
- [x] Make sure that this is on-par with the existing implementation
- [x] Review in detail, now that the large changes are done.
- [ ] Update the roadmap
Release Notes:
- N/A
This PR fixes a panic when attempting to load the `auto_update` setting.
This was leftover from #10296.
I'm going to see if there's a better way we can handle these cases so
they're more obviously correct.
Release Notes:
- N/A
This PR adds the ability for extensions to provide certain language
settings via the language `config.toml`.
These settings are then merged in with the rest of the settings when the
language is loaded from the extension.
The language settings that are available are:
- `tab_size`
- `hard_tabs`
- `soft_wrap`
Additionally, for bundled languages we moved these settings out of the
`settings/default.json` and into their respective `config.toml`s .
For languages currently provided by extensions, we are leaving the
values in the `settings/default.json` temporarily until all released
versions of Zed are able to load these settings from the extension.
---
Along the way we ended up refactoring the `Settings::load` method
slightly, introducing a new `SettingsSources` struct to better convey
where the settings are being loaded from.
This makes it easier to load settings from specific locations/sets of
locations in an explicit way.
Release Notes:
- N/A
---------
Co-authored-by: Max <max@zed.dev>
Co-authored-by: Max Brunsfeld <maxbrunsfeld@gmail.com>
Release Notes:
- Fixed `cgn` backwards movement problem in #9982
There are two issues:
- When there are no more matches, the next repetition still moves the
cursor to the left. After that, the recording is cleared. For this I
simply move the cursor to the right, but it doesn't work when the cursor
is at the end of the line.
- If `cgn` is used when there are no matches, it cleans the previous
recorded actions. Maybe there should be a way to revert the recording.
This also happens when using `c` and `esc`
---------
Co-authored-by: Conrad Irwin <conrad.irwin@gmail.com>
This PR contains various improvements for the markdown preview (some of
which were originally part of #7601).
Some improvements can be seen in the video (see also release notes down
below):
https://github.com/zed-industries/zed/assets/53836821/93324ee8-d366-464a-9728-981eddbfdaf7
Release Notes:
- Added action to open markdown preview in the same pane
- Added support for displaying channel notes in markdown preview
- Added support for displaying the current active editor when opening
markdown preview
- Added support for scrolling the editor to the corresponding block when
double clicking an element in markdown preview
- Improved pane creation handling when opening markdown preview
- Fixed markdown preview displaying non-markdown files
This PR updates the `extension.toml` to allow specifying multiple
languages for a language server to work with.
The `languages` field takes precedence over `language`. In the future
the `language` field will be removed.
As part of this, the Emmet extension has been extended with support for
PHP and ERB.
Release Notes:
- N/A
---------
Co-authored-by: Max <max@zed.dev>
- Fixed#9729 and #10193
This commit fixes an issue where the string splitting function was
handling characters in the input string improperly. We adjusted the use
of the `take_while` function to calculate the length of the numeric
prefix, rather than directly splitting the string, thus correctly
splitting the string into a numeric prefix part and the remaining part
This increases search context from 1 above, 2 below, to 2 above and 2
below, matching the Sublime Text search results.
Release Notes:
- Increased search result context from 3 lines to 4 lines
For #4965
There are still some minor issues:
1. When change the surround and delete the surround, we should also
decide whether there are spaces inside after deleting/replacing
according to whether it is open parentheses, and replace them
accordingly, but at present, delete and change, haven't done this
adaptation for current pr, I'm not sure if I can fit it in the back or
if it needs to be fitted together.
2. In the selection mode, pressing s plus brackets should also trigger
the Add Surrounds function, but this MR has not adapted the selection
mode for the time being, I think we need to support different add
behaviors for the three selection modes.(Currently in select mode, s is
used for Substitute)
3. For the current change surrounds, if the user does not find the
bracket that needs to be matched after entering cs, but it is a valid
bracket, and will wait for the second input before failing, the better
practice here should be to return to normal mode if the first bracket is
not found
4. I reused BracketPair in language, but two of its properties weren't
used in this mr, so I'm not sure if I should create a new struct with
only start and end, which would have less code
I'm not sure which ones need to be changed in the first issue, and which
ones can be revised in the future, and it seems that they can be solved
---------
Co-authored-by: Conrad Irwin <conrad.irwin@gmail.com>
Fixes: #10266
Release Notes:
- Added/Fixed/Improved ...
([#<public_issue_number_if_exists>](https://github.com/zed-industries/zed/issues/<public_issue_number_if_exists>)).
Optionally, include screenshots / media showcasing your addition that
can be included in the release notes.
**or**
- N/A
This PR makes Zed respect the language server's capabilities when
calling the `GetReferences` command (used in "Find All References",
etc.).
This fixes a crash that could occur when using Zed with Gleam v1.0.
Release Notes:
- Made "Find All References" respect the language server's capabilities.
This fixes some instances where certain language servers would stop
working after receiving a "Find All References" request.
---------
Co-authored-by: Max <max@zed.dev>
At the moment, the editor scrollbar is 12px wide. One pixel is allocated
for the left border, so we have 11 pixels to display markers. It's not
enough to make three even marker columns (git, highlights, diagnostics)
that fully fill the scrollbar, so the current implementation allocates 3
pixels to each column.
As the result, we have 2 spare pixels on the right (before #10080 they
were occupied by the diagnostics column). Making the scrollbar just one
pixel wider allows us to give one additional pixel to each marker column
and make markers more pronounced ("as is" on the left, "to be" on the
right):
<img width="115" alt="zed-scrollbar-markers-1px"
src="https://github.com/zed-industries/zed/assets/2101250/4bdf0107-c0f1-4c9c-9063-d2ff461e1c32">
Other options:
- Remove scrollbar thumb border. That'll give us one missing pixel to
make markers wide and even. I, personally, prefer this option, but
themes now have `scrollbar.thumb.border` colors that differ from
`scrollbar.thumb.background` for some reason. This theme setting becomes
deprecated in this case. For the reference: VS Code doesn't have
scrollbar slider borders, IntelliJ IDEA does have them.
- Don't try to make markers evenly wide. For instance, IntelliJ uses
very narrow git-diff markers that are separated from other markers. But
it requires much wider scrollbar (it's 20px in IDEA).
- Use the spare two pixels to make diagnostic markers wider (it's the
pre #10080 approach), or split them between the highlight and diagnostic
markers (have 3px+4px+4px marker columns).
- Do nothing. It leaves us with two unused pixels :(
Release Notes:
- N/A
Related Issues:
- The previous discussion:
https://github.com/zed-industries/zed/pull/9080#issuecomment-1997979968
