This reduces the number of multibuffer syncs from 100,000 to 20,000.
Before this change each editor individually observed the project, so
literally any project change was amplified by the number of editors you
had open.
Now editors listen to their buffers instead of the project, and other
users of `cx.observe` on the project have been updated to use specific
events to reduce churn.
Follow up to #26237
Release Notes:
- Improved performance of Zed in large repos with lots of file system
events.
---------
Co-authored-by: Max Brunsfeld <maxbrunsfeld@gmail.com>
Modified version of #25950. We still use worktree paths, but repo paths
with a status that lie outside the worktree are not excluded; instead,
we relativize them by adding `..`. This makes the list in the git panel
match what you'd get from running `git status` (with the repo's worktree
root as the working directory).
- [x] Implement + test new unrelativization logic
- [x] ~~When collecting repositories, dedup by .git abs path, so
worktrees can share a repo at the project level~~ dedup repos at the
repository selector layer, with repos coming from larger worktrees being
preferred
- [x] Open single-file worktree with diff when activating a path not in
the worktree
Release Notes:
- N/A
Closes#25976
There's a couple states that this covers:
- upon `git init`, no footer is shown at all
- after 1 commit (or when on any parentless commit), the uncommit button
is ~disabled~ hidden
- otherwise commit button is shown
Also updated the button with "meta" tooltip showing human readable
description and git command.
Release Notes:
- N/A
---------
Co-authored-by: Nate Butler <iamnbutler@gmail.com>
Attempt to fix an issue where conflicts from a cherry-pick don't get
cleared out of the git panel after being resolved.
Release Notes:
- Git Beta: Fixed resolution of conflicts from cherry-picks not being
reflected in the git panel
When a worktree is created, we walk up the ancestors of the root path
trying to find a git repository. In particular, if your `$HOME` is a git
repository and you open some subdirectory of `$HOME` that's *not* a git
repository, we end up scanning `$HOME` and everything under it looking
for changed and untracked files, which is often pretty slow. Consistency
here is not very useful and leads to a bad experience.
This PR adds a special case to not consider `$HOME` as a containing git
repository, unless you ask for it by doing the equivalent of `zed ~`.
Release Notes:
- Changed the behavior of git features to not treat `$HOME` as a git
repository unless opened directly
We've had some issues reported with git repositories not getting
detected when they're a strict parent of the worktree root. Add a bit
more logging to understand what's going on here.
Release Notes:
- N/A
Closes https://github.com/zed-industries/zed/issues/25865
Closes https://github.com/zed-industries/zed/pull/25915
In the issue, Zed had caused `.git/lfs/tmp/466102258`-like files to
appear in the directory, which lead to background FS event listener to
handle this as an update, incrementing snapshot's `scan_id`, which lead
to git status rescan, which caused another increment to `status_scan_id`
— incrementing either of the IDs causes the related repo data to be
considered "changed:
41b45eaba7/crates/worktree/src/worktree.rs (L1590-L1605)
hence propagating events to the other parts of the system (e.g. git
blame, which was also active in the issue's case)
```
[2025-03-01T20:01:08+01:00 DEBUG worktree] ignoring event ".git/lfs/tmp/466102258" within unloaded directory
[2025-03-01T20:01:08+01:00 DEBUG worktree] received fs events []
[2025-03-01T20:01:08+01:00 DEBUG worktree] reloading repositories: ["/Users/alex/dev/monorepo/.git"]
[2025-03-01T20:01:08+01:00 DEBUG editor::blame] Status of git repositories updated. Regenerating blame data...
[2025-03-01T20:01:08+01:00 DEBUG editor::blame] Status of git repositories updated. Regenerating blame data...
[2025-03-01T20:01:08+01:00 DEBUG editor::blame] Status of git repositories updated. Regenerating blame data...
```
Due to repo update events sent, another `.git/lfs/tmp/` entry is
created, things start over...
The PR fixes this by ignoring any `.git/lfs/` directory-related FS
events, as needed for the current git status update heuristics.
https://github.com/zed-industries/zed/pull/25915 tried to follow further
and `scan_id` and `status_scan_id` but we do not store all git state in
memory, e.g. head
e0060b92cc/crates/editor/src/editor_tests.rs (L13686)
as
[tests](https://github.com/zed-industries/zed/actions/runs/13631960559/job/38101504549?pr=25915)
show.
Release Notes:
- Improved `.git` scan heuristics
Like the real app, this one infinite loops if you have a diff in an
UnsharedFile.
Release Notes:
- N/A *or* Added/Fixed/Improved ...
---------
Co-authored-by: Max Brunsfeld <maxbrunsfeld@gmail.com>
Done automatically with
> ast-grep -p '$A.background_executor().spawn($B)' -r
'$A.background_spawn($B)' --update-all --globs "\!crates/gpui"
Followed by:
* `cargo fmt`
* Unexpected need to remove some trailing whitespace.
* Manually adding imports of `gpui::{AppContext as _}` which provides
`background_spawn`
* Added `AppContext as _` to existing use of `AppContext`
Release Notes:
- N/A
This PR uses the template merge message in `.git/MERGE_MSG` to populate
the commit message buffer in the git panel. This is done:
- when the commit message buffer is first created
- when the list of merge heads in .git changes, only if the buffer
doesn't already have some text in it
Hopefully this strikes a good balance between convenience and not
stomping on the user's toes.
Release Notes:
- N/A
Follow-up of https://github.com/zed-industries/zed/pull/24307
Brings back the PR and fixes the issue with the git status not
propagated, if computed too slow.
Now, git repo update
* waits in the background for all `scan_dir` repo status updates and
triggers another status update send afterwards
* ensures that the update sent is reported correctly (`scanning = true`)
if either FS or status scan is running still
* during worktree's git statuses updates, bumps `status_scan_id` to
ensure the repo update is reported to all subscribers
Release Notes:
- Improved project panel's speed in large projects
Closes#10232
Context:
We have three ways to open files or dirs in Zed: `zed`, `zed --new`, and
`zed --add`. `--new` forces the project to open in a new window, while
`--add` forces it to open in an existing window (even if the dir isn’t a
subdir of an existing project or the file isn’t part of it).
Using just `zed` tries to open it in an existing window based on similar
logic of `--add`, but if no related project is found the dir, opens in a
new window.
Problem:
Right now, subdirs that are part of an existing project open in the
existing window when using `zed`. By default, subdirs should open in a
new window instead. If someone wants to open it in the existing window,
they can explicitly use `--add`. After this PR, only root dir and files
will focus on existing window, when `zed ` is used.
Fix:
For the `zed` case, we’ve filtered out subdirs in the logic that assigns
them to an existing window.
Release Notes:
- Fixed an issue where subdirectories of an already opened project, when
opened via the terminal, would open in the existing project instead of a
new window.
Closes#7923
This PR fixes root worktree renaming by:
1. Handling the case where `new_path` is the new root name instead of a
relative path from the root.
2. [#20313](https://github.com/zed-industries/zed/pull/20313) added
functionality to watch for root worktree renames made externally, e.g.,
via Finder. This PR avoids relying on that watcher because, when
renaming explicitly from Zed, we can eagerly perform the necessary work
(of course after fs rename) instead of waiting for the watcher to detect
the rename. This prevents UI glitches during renaming root.
Todo:
- [x] Fix wrong abs paths when root is renamed
- [x] Fix explicit scan entry func to handle renamed root dir
- [x] Tests
- [x] Test on Linux
- [x] Tested with single and multipe worktrees
- [x] Tested when single file is root file
Release Notes:
- Fixed an issue where worktree root name couldn't be renamed in project
panel.
This PR adds the branch selector to the git panel and fixes a few bugs
in the repository selector.
Release Notes:
- N/A
---------
Co-authored-by: ConradIrwin <conrad.irwin@gmail.com>
Co-authored-by: Conrad <conrad@zed.dev>
This is a refactoring PR that does three things:
- First, it introduces a new `diff` crate that holds the previous
contents of the `git::diff` module, plus the `BufferChangeSet` type
formerly of `project::buffer_store`. The new crate is necessary since
simply moving `BufferChangeSet` into `git::diff` results in a dependency
cycle due to the use of `language::Buffer` to represent the diff base in
`BufferChangeSet`.
- Second, it renames the two main types in the new diff crate:
`BufferDiff` becomes `BufferDiffSnapshot`, and `BufferChangeSet` becomes
`BufferDiff`. This reflects that the relationship between these two
types (immutable cheaply-cloneable "value" type + stateful "resource
type" with subscriptions) mirrors existing pairs like
`Buffer`/`BufferSnapshot`. References to "change sets" throughout the
codebase are updated to refer to "diffs" instead.
- Finally, it moves the base_text field of the new BufferDiff type to
BufferDiffSnapshot.
Release Notes:
- N/A
---------
Co-authored-by: maxbrunsfeld <max@zed.dev>
Go back to a less optimized implementation for now since the custom
cursor target seems to have some bugs.
Release Notes:
- Fixed missing git blame and status output in some projects with
multiple git repositories
Rework of https://github.com/zed-industries/zed/pull/24130
Uses
1033c0b57e
`COMMIT_EDITMSG` language-related definitions (thanks @d1y )
Instead of using real `.git/COMMIT_EDITMSG` file, create a buffer
without FS representation, stored in the `Repository` and shared the
regular way via the `BufferStore`.
Adds a knowledge of what `Git Commit` language is, and uses it in the
buffers which are rendered in the git panel.
Release Notes:
- N/A
---------
Co-authored-by: Conrad Irwin <conrad@zed.dev>
Co-authored-by: d1y <chenhonzhou@gmail.com>
Co-authored-by: Smit <smit@zed.dev>
This PR builds on #21258 to make it possible to use HEAD as a diff base.
The buffer store is extended to support holding multiple change sets,
and collab gains support for synchronizing the committed text of files
when any collaborator requires it.
Not implemented in this PR:
- Exposing the diff from HEAD to the user
- Decorating the diff from HEAD with information about which hunks are
staged
`test_random_multibuffer` now fails first at `SEED=13277`, similar to
the previous high-water mark, but with various bugs in the multibuffer
logic now shaken out.
Release Notes:
- N/A
---------
Co-authored-by: Max <max@zed.dev>
Co-authored-by: Ben <ben@zed.dev>
Co-authored-by: Max Brunsfeld <maxbrunsfeld@gmail.com>
Co-authored-by: Conrad Irwin <conrad.irwin@gmail.com>
Co-authored-by: Conrad <conrad@zed.dev>
A preparation for collaborative commit message editing.
Before, almost any `.git`-contained file FS update, except
`.git/fsmonitor--daemon/cookies/**` caused git metadata rescan.
This included `index.lock` that was created after any git operation,
e.g. `git status`, which was unnecessary.
Collaborative editing aims to share `.git/COMMIT_EDITMSG` between
multiple users, so there are potentially multiple users editing the file
and causing excessive events.
The change makes worktree to ignore .git/COMMIT_EDITMSG`,
`.git/index.lock` and `.git/fsmonitor--daemon/**` paths and adjusts the
logic to be more extensible: there's much more files Zed can ignore and
still have its git metadata up to date.
Release Notes:
- N/A
In both `thread` and `prompt editor` the context file picker, gitignored
files are hidden (as expected) when searching files by path, but they
are still shown initially as you create the file picker.
Plus, selecting gitignored files in the `prompt editor` is bugged and
collapses everything.
This PR settles on not showing gitignored files to solve these
inconsistencies.
Release Notes:
- Fix gitignored files filter occasionally not working in context file
picker.
Closes#15966
This PR adds `Alt/Opt+Click` to expand or collapse a directory and all
its contents.
Context:
The current `expand_entry` scans immediate child subdirectories if they
aren’t loaded, while `expand_all_for_entry` scans the entire subtree.
The latter takes longer, so we wait for it to complete to ensure
accurate results.
For full directory scan, instead of using
`refresh_entries_for_paths(vec![path])`, which requires specifying all
explicit paths to refresh, we use `add_path_prefix_to_scan`, which
eliminates the need to list every path. Both methods internally call
`reload_entries_for_paths`, which invokes `should_scan_directory`. This
determines whether to scan deeper based on a path prefix match between
the given directory and its subdirectories, returning `true` for
`add_path_prefix_to_scan`.
The existing code handles scanning, removing path prefixes after scans
complete, and managing ignored directories.
How it works (Expand):
1. Alt clicking on non-ignored closed directory, expands it and all its
subdirectories, except ignored subdirectories. This helps while working
on mono repos, where you might not want to expand dirs like
`node_modules`, `dist`, etc or git submodules, when you expand any root
dir.
In example, `draft` and `posts` dir are ignored dir.
[expand-1.webm](https://github.com/user-attachments/assets/07d3f724-0757-408f-b349-5beb4ee8440e)
2. Alt clicking on ignored closed directory, expands it and all its
subdirectories. This is when you explicitly want to do it, on dirs like
`node_modules`, `dist`, etc.
In example, `dist` dir is ignored dir.
[expand-2.webm](https://github.com/user-attachments/assets/99e55883-ab1a-4a9c-a0f0-48026991a922)
3. In case of auto folded subdirectories, expand all action will take
precedence over it. That is, it will unfold all the subdirectories
inside clicked dir. This is intentional, as user explicitly wants to
reveal as much content as possible. (This is my personal opinion on how
it should work).
[expand-3.webm](https://github.com/user-attachments/assets/f20b0311-e92a-4e34-b640-1469b0d6fa16)
How it works (Collapse):
1. Alt clicking any opened directory will collapse it and all its
children, whether ignored or not. This is when you want to start from a
fresh state.
2. When auto fold is enabled in settings, collapse action will also fold
all subdirectories that it can fold. This is to bring it back to its
fresh state as mentioned above.
[collapse-1-2.webm](https://github.com/user-attachments/assets/74db6cee-0afa-406b-a9a2-7421083a2c2a)
Future:
- Using keybinding to expand/collapse all for selected entry
- Handle expand/collapse all for folded entry
Todos:
- [x] Expand entries logic
- [x] Handle remote worktree for expand
- [x] Figure out scan complete status
- [x] Move expansion logic to status update event
- [x] Collapse entries logic
- [x] Handle fold/unfold subdirs interaction
- [x] Do not expand git ignored sub-dirs
- [x] Tests
- [x] Test Remote
Release Notes:
- Added Alt/Opt+Click functionality to expand or collapse a directory
and all its contents.
Fix a bug where a GPUI macro still used `ModelContext`
Rename `AsyncAppContext` -> `AsyncApp`
Rename update_model, read_model, insert_model, and reserve_model to
update_entity, read_entity, insert_entity, and reserve_entity
Release Notes:
- N/A
In the Windows test environment, the paths generated by `temp_tree()`
are symlink paths, which causes certain tests to fail.
I later noticed that when opening a project, we seem to always use
`canonicalize` to normalize the paths, as shown here:
https://github.com/zed-industries/zed/pull/21039.
This PR adopts a similar approach for the test environment to address
the issue.
Release Notes:
- N/A
There's still a bit more work to do on this, but this PR is compiling
(with warnings) after eliminating the key types. When the tasks below
are complete, this will be the new narrative for GPUI:
- `Entity<T>` - This replaces `View<T>`/`Model<T>`. It represents a unit
of state, and if `T` implements `Render`, then `Entity<T>` implements
`Element`.
- `&mut App` This replaces `AppContext` and represents the app.
- `&mut Context<T>` This replaces `ModelContext` and derefs to `App`. It
is provided by the framework when updating an entity.
- `&mut Window` Broken out of `&mut WindowContext` which no longer
exists. Every method that once took `&mut WindowContext` now takes `&mut
Window, &mut App` and every method that took `&mut ViewContext<T>` now
takes `&mut Window, &mut Context<T>`
Not pictured here are the two other failed attempts. It's been quite a
month!
Tasks:
- [x] Remove `View`, `ViewContext`, `WindowContext` and thread through
`Window`
- [x] [@cole-miller @mikayla-maki] Redraw window when entities change
- [x] [@cole-miller @mikayla-maki] Get examples and Zed running
- [x] [@cole-miller @mikayla-maki] Fix Zed rendering
- [x] [@mikayla-maki] Fix todo! macros and comments
- [x] Fix a bug where the editor would not be redrawn because of view
caching
- [x] remove publicness window.notify() and replace with
`AppContext::notify`
- [x] remove `observe_new_window_models`, replace with
`observe_new_models` with an optional window
- [x] Fix a bug where the project panel would not be redrawn because of
the wrong refresh() call being used
- [x] Fix the tests
- [x] Fix warnings by eliminating `Window` params or using `_`
- [x] Fix conflicts
- [x] Simplify generic code where possible
- [x] Rename types
- [ ] Update docs
### issues post merge
- [x] Issues switching between normal and insert mode
- [x] Assistant re-rendering failure
- [x] Vim test failures
- [x] Mac build issue
Release Notes:
- N/A
---------
Co-authored-by: Antonio Scandurra <me@as-cii.com>
Co-authored-by: Cole Miller <cole@zed.dev>
Co-authored-by: Mikayla <mikayla@zed.dev>
Co-authored-by: Joseph <joseph@zed.dev>
Co-authored-by: max <max@zed.dev>
Co-authored-by: Michael Sloan <michael@zed.dev>
Co-authored-by: Mikayla Maki <mikaylamaki@Mikaylas-MacBook-Pro.local>
Co-authored-by: Mikayla <mikayla.c.maki@gmail.com>
Co-authored-by: joão <joao@zed.dev>
