This PR improves workflow step management and symbol matching. We've
optimized step pruning to remove any step that intersects an edit and
switched to normalized Levenshtein distance for more accurate symbol
matching.
Release Notes:
- N/A
This PR opens workflow step editors as preview tabs and closes them upon
exiting the step if they are still in preview mode and they weren't
already open before entering the step.
Making this work was tricky, because we often edit the buffer as part of
displaying the workflow step suggestions to create empty lines where we
can generate. We undo these edits if the transformation is not applied,
but they were causing the preview to be dismissed.
After trying a few approaches, I decided to give workspace `Item`s a
`preserve_preview` method that defaults to false. When the workspace
sees an edit event for the item, it checks if the item wants to preserve
its preview. For buffers, after editing, you can call `refresh_preview`,
which sets a preview version to the current version of the buffer. Any
edits after this version will cause preview to not be preserved.
One final issue is with async auto-indent. To ensure these async edits
don't dismiss the preview, I automatically refresh the preview version
if preview was preserved prior to performing the auto-indent. The
assumption is that these are edits created by other edits, and if we
didn't want to dismiss the preview with the originating edits, then the
auto-indent edits shouldn't dismiss it either.
Release Notes:
- N/A
---------
Co-authored-by: Jason <jason@zed.dev>
This commit weaves through new APIs for language::BufferChunks, multi_buffer::MultiBufferChunks and inlay_map::InlayChunks that allow seeking with an upper-bound. This allows us to omit doing syntax highligting and looking up diagnostics for folded ranges. This in turn directly improves performance of assistant panel with large contexts.
Release Notes:
- Fixed poor performance when editing in the assistant panel after
inserting large files using slash commands
---------
Co-authored-by: Max <max@zed.dev>
Co-authored-by: Max Brunsfeld <maxbrunsfeld@gmail.com>
This PR removes the primary/secondary distinction for
`CachedLspAdapter`s.
After #15624 we weren't relying on the `is_primary` field anywhere, so
we can remove it.
Release Notes:
- N/A
This PR updates how we determine the "primary" language server for a
buffer to make it respect the order specified by the `language_servers`
setting.
Previously we were relying on the language servers to be registered in
the right order in order to select the primary one effectively.
However, in my testing I observed some cases where a native language
server (e.g., `tailwindcss-language-server`) could end up first in the
list of language servers despite not being first in the
`language_servers` setting.
While this wasn't a problem for the Tailwind or ESLint language servers
on account of them being defined natively with the designation of
"secondary" language servers, this could cause problems with
extension-based language servers.
To remedy this, every time we start up language servers we reorder the
list of language servers for a given language to reflect the order in
the `language_servers` setting. This ordering then allows us to treat
the first language server in the list as the "primary" one.
Related issues:
- https://github.com/zed-industries/zed/issues/15023
- https://github.com/zed-industries/zed/issues/15279
Release Notes:
- The ordering of language servers will now respect the order in the
`language_servers` setting.
- The first language server in this list will be used as the primary
language server.
This PR extends the fix from #15336 to more places that had the same
issue.
An `add_references_to_properties` helper function has been added to
handle these cases uniformly.
Release Notes:
- N/A
Supersedes https://github.com/zed-industries/zed/pull/12090fixes#5180fixes#5055
See original PR for an example of the feature at work.
This PR changes the settings interface to be backwards compatible, and
adds the `ui_font_fallbacks`, `buffer_font_fallbacks`, and
`terminal.font_fallbacks` settings.
Release Notes:
- Added support for font fallbacks via three new settings:
`ui_font_fallbacks`, `buffer_font_fallbacks`, and
`terminal.font_fallbacks`.(#5180, #5055).
---------
Co-authored-by: Junkui Zhang <364772080@qq.com>
Closes https://github.com/zed-industries/zed/issues/4565
To fix issues with code blocks' parsing in Markdown, a
tree-sitter-markdown library update is needed.
But `tree_sitter::language` is used in many places within core Zed,
which forced more library updates.
Release Notes:
- Updated tree-sitter parsers for core languages
---------
Co-authored-by: Max Brunsfeld <max@zed.dev>
Co-authored-by: Piotr Osiewicz <piotr@zed.dev>
Fixes#4822
- [x] Release note
- [ ] Surface formatting errors via a toast
- [x] Doc updates
- [x] Have "language-server" accept an optional name of the server.
Release Notes:
- `format` and `format_on_save` now accept an array of formatting
actions to run.
- `language_server` formatter option now accepts the name of a language
server to use (e.g. `{"language_server": {"name: "ruff"}}`); when not
specified, a primary language server is used.
---------
Co-authored-by: Thorsten <thorsten@zed.dev>
Per the LSP spec, we should pass .data field of diagnostics into code
action request:
```
/**
* A data entry field that is preserved between a
* `textDocument/publishDiagnostics` notification and
* `textDocument/codeAction` request. *
* @since 3.16.0 */ data?: LSPAny;
```
Release Notes:
- Fixed rare cases where a code action triggered by diagnostic may not
be available for use.
This PR updates a number of spots where we were setting all of the
`TextStyle` fields even if we were not changing the values from the
defaults.
We now use `..Default::default()`.
Release Notes:
- N/A
This is a first step towards allowing you to edit remote projects
directly over SSH. We'll start with a pretty bare-bones feature set, and
incrementally add further features.
### Todo
Distribution
* [x] Build nightly releases of `zed-remote-server` binaries
* [x] linux (arm + x86)
* [x] mac (arm + x86)
* [x] Build stable + preview releases of `zed-remote-server`
* [x] download and cache remote server binaries as needed when opening
ssh project
* [x] ensure server has the latest version of the binary
Auth
* [x] allow specifying password at the command line
* [x] auth via ssh keys
* [x] UI password prompt
Features
* [x] upload remote server binary to server automatically
* [x] opening directories
* [x] tracking file system updates
* [x] opening, editing, saving buffers
* [ ] file operations (rename, delete, create)
* [ ] git diffs
* [ ] project search
Release Notes:
- N/A
---------
Co-authored-by: Piotr Osiewicz <24362066+osiewicz@users.noreply.github.com>
Note that this shouldn't have any visible user-facing behavior yet. The
feature is incomplete but we wanna merge early to avoid a long-running
branch.
Release Notes:
- N/A
---------
Co-authored-by: Nathan <nathan@zed.dev>
The biggest hurdle turned out to be use of `Arc<Language>` in maps, as
`clippy::mutable_key_type` started triggering on it (due to - I suppose
- internal mutability on `HighlightMap`?). I switched over to using
`LanguageId` as the key type in some of the callsites, as that's what
`Language` uses anyways for it's hash/eq, though I've still had to
suppress the lint outside of language crate.
/cc @maxdeviant , le clippy guru.
Release Notes:
- N/A
This pull request introduces collaboration for the assistant panel by
turning `Context` into a CRDT. `ContextStore` is responsible for sending
and applying operations, as well as synchronizing missed changes while
the connection was lost.
Contexts are shared on a per-project basis, and only the host can share
them for now. Shared contexts can be accessed via the `History` tab in
the assistant panel.
<img width="1819" alt="image"
src="https://github.com/zed-industries/zed/assets/482957/c7ae46d2-cde3-4b03-b74a-6e9b1555c154">
Please note that this doesn't implement following yet, which is
scheduled for a subsequent pull request.
Release Notes:
- N/A
Context:
@bennetbo spotted a regression in handling of `cargo run` task in zed
repo following a merge of #13658. We've started invoking `cargo run`
from the folder of an active file whereas previously we did it from the
workspace root. We brainstormed few solutions that involved adding a
separate task that gets invoked at a workspace level, but I realized
that a cleaner solution may be to finally add user-configured task
variables. This way, we can choose which crate to run by default at a
workspace level.
This has been originally brought up in the context of javascript tasks
in
https://github.com/zed-industries/zed/pull/12118#issuecomment-2129232114
Note that this is intended for internal use only for the time being.
/cc @RemcoSmitsDev we should be unblocked on having runner-dependant
tasks now.
Release notes:
- N/A
As a drive-by of https://github.com/zed-industries/zed/pull/13654, I've
noticed that the editor felt sluggish after I've undone the changes made
by the replacement. It turns out that we are repeatedly checking whether
there are any edits to estabilish dirty/conflict state of a buffer, even
though this operation is pure; this PR stores away the result of a
computation and refers to it before rerunning it.
Release Notes:
- Improve editor's performance with large undo histories
Buffers carry several pieces of state besides their text: syntax tree,
diagnostics, git diff, and file data. Previously, the buffer maintained
a separate integer version number for each of these four pieces of
state, incrementing it every time that piece of state is updated. This
is used by MultiBuffers to detect when they need to update excerpts.
Previously, for a given buffer, these four version numbers were stored
on the buffer itself, on every snapshot of the buffer, in any
multi-buffer that referenced that buffer, **and** on snapshots of that
multi-buffer. But the only use for the version numbers was reduced down
to a single boolean predicate: whether or not the buffer's state has
changed.
In this PR, I've combined those 4 version numbers into one. I've called
it `non_text_state_update_count` because it tracks all state updates
outside of the text itself. This removes a bunch of unnecessary code,
and reduces the size of buffer snapshots and multi-buffer snapshots.
Release Notes:
- N/A
In the past, Zed used a single switch called `autoclose` to control both
`autoclose` and `auto_surround` functionalities:
+ `autoclose`: when input '(', append ')' automatically.
+ `auto_surround`: when select text and input '(', surround text with
'(' and ')' automatically.
This PR separates `auto_surround` from `autoclose` to support `<`.
Previously, if `autoclose` of `<` was set to `false`, `auto_surround`
couldn't be used. However, setting `autoclose` to `true` would affect
the default behavior of simple expression. For example, `a < b` would
become `a <> b`.
For more information, see #13187.
Fix#12898.
Release Notes:
- Added support for `auto_surround`
([#12898](https://github.com/zed-industries/zed/issues/12898)).
This PR removes the Copilot-specific aliases for the
`inline_completions` and `show_inline_completions` settings.
While these aliases were added to maintain backward-compatibility, the
aliasing behavior here can lead to a confusing experience when both keys
end up in the `settings.json`.
Release Notes:
- Breaking Change: Removed the `copilot` alias for the
`inline_completions` setting. If you have settings under `copilot` they
should get moved to `inline_completions`.
- Breaking Change: Removed the `show_copilot_suggestions` alias for the
`show_inline_completions` setting.
This PR removes the `language_overrides` alias for the `languages`
setting.
I've seen a number of people run into issues where they have both
`languages` and `language_overrides` in their settings and get confused
when their settings don't seem to apply as expected.
This is a breaking change, but I think it is a necessary one to prevent
more users from running into issues.
Release Notes:
- Breaking Change: Removed the `language_overrides` alias for the
`languages` setting. If you have settings under `language_overrides`
they should get moved to `languages`.
This PR adds support for [linked editing of
ranges](https://microsoft.github.io/language-server-protocol/specifications/lsp/3.17/specification/#textDocument_linkedEditingRange),
which in short means that editing one part of a file can now change
related parts in that same file. Think of automatically renaming
HTML/TSX closing tags when the opening one is changed.
TODO:
- [x] proto changes
- [x] Allow disabling linked editing ranges on a per language basis.
Fixes#4535
Release Notes:
- Added support for linked editing ranges LSP request. Editing opening
tags in HTML/TSX files (with vtsls) performs the same edit on the
closing tag as well (and vice versa). It can be turned off on a language-by-language basis with the following setting:
```
"languages": {
"HTML": {
"linked_edits": true
},
}
```
---------
Co-authored-by: Bennet <bennet@zed.dev>
The `worktree` crate mainly provides an in-memory model of a directory
and its git repositories. But because it was originally extracted from
the Project crate, it also contained lingering bits of code that were
outside of that area:
* it had a little bit of logic related to buffers (though most buffer
management lives in `project`)
* it had a *little* bit of logic for storing diagnostics (though the
vast majority of LSP and diagnostic logic lives in `project`)
* it had a little bit of logic for sending RPC message (though the
*receiving* logic for those RPC messages lived in `project`)
In this PR, I've moved those concerns entirely to the project crate
(where they were already dealt with for the most part), so that the
worktree crate can be more focused on its main job, and have fewer
dependencies.
Worktree no longer depends on `client` or `lsp`. It still depends on
`language`, but only because of `impl language::File for
worktree::File`.
Release Notes:
- N/A
Indent guides can be configured per language, meaning that in a multi
buffer we can get excerpts where indent guides should be
disabled/enabled/styled differently than other excerpts.
Imagine the following scenario, i have indent guides disabled in my
settings, but want to enable them for JS and Python. I also want to use
a different line width for python files. Something like this is now
supported:
<img width="445" alt="image"
src="https://github.com/zed-industries/zed/assets/53836821/0c91411c-145c-4210-a883-4c469d5cb828">
And the relevant settings for the example above:
```json
"indent_guides": {
"enabled": false
},
"languages": {
"JavaScript": {
"indent_guides": {
"enabled": true
}
},
"Python": {
"indent_guides": {
"enabled": true,
"line_width": 5
}
}
}
```
Release Notes:
- Respect language specific settings when showing indent guides in a
multibuffer
- Fixes an issue where indent guide specific settings were not
recognized when specified in local settings
When indent guides were still WIP, I thought it might be a good idea to
detect the tab size for every line individually, so we can handle files
with mixed indentations. However, while optimizing the performance of
indent guides I found that getting the language at a given anchor was
pretty expensive, therefore I only resolved the language for the first
visible row. However, this could lead to some weird flickering, where
the indent guides would use different tab sizes depending on the first
visible row (see #12492). This can be fixed by just using the primary
buffer language size.
So as of right now indent guides cannot handle files with mixed
indentations. Im not sure if anyone actually does/expects this, but one
use case I could imagine is something like this:
User x has a svelte file, where the tab size is set to `4`. However the
svelte code uses typescript inside a script tag, which User x wants to
use a tab size of `2`. The approach used here would not work for this,
but then again I think our formatter does not even support something
like this. Im probably overcomplicating things, so let's stick with the
simple solution for now.
Release Notes:
- Fixed an issue where indent guides would use an incorrect tab size
([#12492](https://github.com/zed-industries/zed/issues/12492)).
Follow-up of
https://github.com/zed-industries/zed/pull/12095#issuecomment-2123230762
reverting back part of https://github.com/zed-industries/zed/pull/11558
that was related to `language.toml` parsing.
Now all extensions that define `prettier_parser_name` in their language
configs, will enable formatting untitled buffers without any extra
language settings like
```json
{
"languages": {
"JSON": {
"prettier": {
"allowed": true,
"parser": "json"
}
}
}
}
```
Release Notes:
- Improved ergonomics of untitled buffer formatting with prettier, no
extra language settings are needed by default.
In #12003 we found ourselves in need for precise region tracking in
which a given runnable has an effect in order to grab variables from it.
This PR makes it so that in task modal all task variables from queries
overlapping current cursor position.
However, in the process of working on that I've found that we cannot
always use a top-level capture to represent the full match range of
runnable (which has been my assumption up to this point). Tree-sitter
captures cannot capture sibling groups; we did just that in Rust
queries.
Thankfully, none of the extensions are affected as in them, a capture is
always attached to single node. This PR adds annotations to them
nonetheless; we'll be able to get rid of top-level captures in extension
runnables.scm once this PR is in stable version of Zed.
Release Notes:
- N/A
This pull request replaces the static `⋯` character we used to insert
when folding a range with a custom render function that return an
`AnyElement`. We plan to use this in the assistant, but for now this
should be behavior-preserving.
Release Notes:
- N/A
---------
Co-authored-by: Nathan <nathan@zed.dev>
Co-authored-by: Conrad <conrad@zed.dev>
