Closes #ISSUE
## background
If a project is big, some times it will be splited into many small git
repos.
[google repo](https://gerrit.googlesource.com/git-repo/) is a tool to
manage a group of git repos.
But, any small git repo manged by this tool, have a difference with
normal git repo.
That is , the path `.git` in the root of the git repo, is not a normal
directory, but a soft link to real git bare dir.
### zed can not recognize the `git-repo` managed git repos
you can use the procedure to genreate this problem
```bash
# tested on linux
mkdir -p bad_git_repo_project
cd bad_git_repo_project
git init
echo "hello" > hi.txt
git add .
git commit -m "init commit"
echo "hello world" >> hi.txt
# modify the repo
mv .git ../.real_git_repo
ln -sf ../.real_git_repo .git
```
with vscode, after opening this project, git works well.
but for Zed, git not work(not git status, no git blame)
## how to fix
libgit2 can recognize git repo from the root of the project(dir that
have `.git`).
so, we can recognize the git project by opening from the project root
dir, but not the `.git` dir
This fix also works with normal git project.
### before fix
### after fix
Release Notes:
- Fix opening repos when .git is a soft link
This change allows defining keybindings that are active when there is a
text selection.
This is especially useful, as an example, for Emacs-like keybindings
where movement keybindings expand the selection.
Here is a snippet from my keymap.json that implements Emacs movements
when selection is active:
```json
{
"context": "Editor && selection",
"bindings": {
"ctrl-f": "editor::SelectRight",
"ctrl-b": "editor::SelectLeft",
"ctrl-n": "editor::SelectDown",
"ctrl-p": "editor::SelectUp",
"ctrl-a": "editor::SelectToBeginningOfLine",
"ctrl-e": "editor::SelectToEndOfLine",
"alt-f": "editor::SelectToNextWordEnd",
"alt-b": "editor::SelectToPreviousWordStart",
"alt-<": "editor::SelectToBeginning",
"alt->": "editor::SelectToEnd"
}
}
```
What do you think about inclusion of this feature? Should I add more granular `selection=single` `selection=multi`?
Release Notes:
- Added "selection" context for keybindings that are active when there is a text selection.
Closes#11115
**Context**:
Consider a monorepo setup like this: the root has Prettier installed,
but the individual monorepos do not. In this case, only one Prettier
instance is used, with its installation located at the root. The
monorepos also use this same instance for formatting.
However, monorepo can have its own `.prettierignore` file, which will
take precedence over the `.prettierignore` file at the root level (if
one exists) for files in that monorepo.
<img
src="https://github.com/user-attachments/assets/742f16ac-11ad-4d2f-a5a2-696e47a617b9"
alt="prettier" width="200px" />
**Implementation**:
From the context above, we should keep ignore dir decoupled from the
Prettier instance. This means that even if the project has only one
Prettier installation (and thus a single Prettier instance), there can
still be multiple `.prettierignore` in play.
This approach also allows us to respect `.prettierignore` even when the
project does not have Prettier installed locally and instead relies on
the editor’s Prettier instance.
**Tests**:
1. No Prettier in project, using editor Prettier: Ensures
`.prettierignore` is respected even without a local Prettier
installation.
2. Monorepo with root Prettier and child `.prettierignore`: Confirms
that the child project’s ignore file is correctly used.
3. Monorepo with root and child `.prettierignore` files: Verifies the
child ignore file takes precedence over the root’s.
Release Notes:
- Added `.prettierignore` support to the Prettier integration.
Closes https://github.com/zed-industries/zed/issues/6843
I don't see where is the logic to remove old document highlight when new
one applies,
ideally, old highlight should be cleared as soon as possible when cursor
moves if the new position does not
sits in old highlight ranges to avoid linger highlights described in
https://github.com/zed-industries/zed/issues/13682#issuecomment-2498368680.
So current solution is still not ideal, because only when lsp responses
highlight ranges (even is a empty set) can we clear the old one.
Release Notes:
- Added a setting `lsp_highlight_debounce` to configure delay for
querying highlights from language server.
---------
Co-authored-by: mgsloan@gmail.com <michael@zed.dev>
Earlier work by @osiewicz in #21931 aims to fix audio control tooltips
in the title bar to close#21929. However, its logic is not quite
correct, and does not match the toggle behavior for the controls.
This PR corrects the logic to match the toggle behavior for the
controls. It also updates capitalization and wording for consistency.
Release Notes:
- N/A
While working on Assistant2, I noticed that the `LanguageModelSelector`
was recreating its `Picker` view on every single render.
This PR makes it so we create the view once and hold onto it in the
parent view.
Release Notes:
- N/A
Some VPN configurations require that websockets present a user agent.
This adds it in directly for the repl usage. I wish there was a way to
reuse the user agent from the `cx.http_client`, but I'm not seeing a
simple way to do that for the moment since it's not on the `HttpClient`
trait.
No release notes since this feature hasn't been announced/exposed.
Release Notes:
- N/A
- Add and wire through git method stubs
- Organize render methods
- Track modifier changes
- Swap commit buttons when `option`/`alt` is held
- More TODOs
Release Notes:
- N/A
`LanguageModel::stream_completion_text` can poll the `stream_completion`
stream (ultimately a `futures::Unfold`) after it's returned
`Ready(None)`, which leads to a panic; avoid this by fusing the stream.
Release Notes:
- Fixed a panic when streaming language model completions
Improve behavior of ctrl-a/ctrl-e home/end in emacs keybind.
Follow up to #21921 to add those to Linux emacs keymap too.
Release Notes:
- emacs: Improved `ctrl-a` / `ctrl-e` / `home` / `end` behavior
- emacs: Added for `ctrl-s` / `ctrl-r` / `ctrl-g` for navigating buffer
search results
When a user types, chances are the model will anticipate what they are
about to do. Previously, we would continuously cancel the pending
completion until the user stopped typing. With this commit, we allow at
most two completions to race with each other (the first and the last
one):
- If the completion that was requested first completes first, we will
show it (assuming we can interpolate it) but avoid canceling the last
one.
- When the completion that was requested last completes, we will cancel
the first one if it's pending.
In both cases, if a completion is already on-screen we have a special
case for when the completions are just insertions and the new completion
is a superset of the existing one. In this case, we will replace the
existing completion with the new one. Otherwise we will keep showing the
old one to avoid thrashing the UI.
This should make latency a lot better. Note that I also reduced the
debounce timeout to 8ms.
Release Notes:
- N/A
Most notably, adding a current iteration of a possible logo to feel it
out! :) Also, I'm hiding the input and instructions container after the
review has been sent. In the future, if we allow changing an already
sent review, we can change this behavior.
<img width="800" alt="Screenshot 2024-12-12 at 10 42 44"
src="https://github.com/user-attachments/assets/37e63d0d-d847-445e-bdf8-bf5c97d0fe4c"
/>
Release Notes:
- N/A
The Linux watcher was unconditionally watching the parent directory of
every watched path. This is needed in the case of config files that may
not exist when the watch is set up, but not in other cases. Scoping the
parent watch more narrowly cuts down on the amount of error logging from
irrelevant file change notifications being sent to Zed (in my case it
was picking up changes to a random file in `$HOME`).
Release Notes:
- N/A
Adds a simple git placeholder panel for us to iterate from. Also
includes a number of assets from the git prototyping branch that we will
use.
Note: This panel is staff flagged for now.
Release Notes:
- N/A
Still TODO:
* [x] Factor out `start_language_server` so we can call it on register
(instead of on detect language)
* [x] Only call register in singleton editors (or when
editing/go-to-definition etc. in a multibuffer?)
* [x] Refcount on register so we can unregister when no buffer remain
* [ ] (maybe) Stop language servers that are no longer needed after some
time
Release Notes:
- Fixed language servers starting when doing project search
- Fixed high CPU usage when ignoring warnings in the diagnostics view
---------
Co-authored-by: Piotr Osiewicz <24362066+osiewicz@users.noreply.github.com>
Co-authored-by: Cole <cole@zed.dev>
This is behind the Assistant v2 feature flag. As @maxdeviant and I
discussed, the state is currently decoupled from the Assistant Panel's
state, although in the future we plan to introduce a way to refer to
conversations from the panel. Also, we're intentionally duplicating some
code with the v2 panel right now; the plan is to do a future PR to make
them share code more.
https://github.com/user-attachments/assets/bb163bd3-a02d-4a91-8f8f-2a8e60acbc34
It doesn't include the terminal inline assistant, which will be in a
separate PR.
Release Notes:
- N/A
This PR updates the `gpui::prelude` to not export the `Context` trait
named.
This prevents some naming clashes in downstream consumers.
Release Notes:
- N/A
This changes the behavior of how we display inline completions and
non-inline completions (i.e. completion menu).
Previously we would never show inline completions if a completion menu
was visible, meaning that we'd never show Copilot/Supermaven/...
suggestions if the language server had a suggestion.
With this change, we now display the inline completions even if there is
a completion menu visible.
In that case `<tab>` then accepts the inline completion and `<enter>`
accepts the selected entry in the completion menu.
Release Notes:
- Changed how inline completions (Copilot, Supermaven, ...) and normal
completions (from language servers) interact. Zed will now also show
inline completions when the completion menu is visible. The user can
accept the inline completion with `<tab>` and the active entry in the
completion menu with `<enter>`. Previously, `<tab>` would also select
the active entry in the completion menu.
---------
Co-authored-by: Antonio <antonio@zed.dev>
If the instructions are added as the input placeholder, when in a
smaller window size (like the one from the screenshot), scrolling is
needed to see them all. So, thought of extracting it out of there. Also
thought it looked more refined this way!
<img width="800" alt="Screenshot 2024-12-11 at 11 48 17"
src="https://github.com/user-attachments/assets/46974b94-6365-4a59-bf71-a6c0863aac68"
/>
Release Notes:
- N/A
Nate Butler