## Context
While looking into: #32051 and #16120 with instruments, I noticed that
`TabSnapshot::to_tab_point` and `TabSnapshot::to_fold_point` are a
common bottleneck between the two issues. This PR takes the first steps
into closing the stated issues by improving the performance of both
those functions.
### Method
`to_tab_point` and `to_fold_point` iterate through each character in
their rows to find tab characters and translate those characters into
their respective transformations. This PR changes this iteration to take
advantage of the tab character bitmap in the `Rope` data structure and
goes directly to each tab character when iterating.
The tab bitmap is now passed from each layer in-between the `Rope` to
the `TabMap`.
### Testing
I added several randomized tests to ensure that the new `to_tab_point`
and `to_fold_point` functions have the same behavior as the old methods
they're replacing. I also added `test_random_chunk_bitmap` on each layer
the tab bitmap is passed up to the `TabMap` to make sure that the bitmap
being passed is transformed correctly between the layers of
`DisplayMap`.
`test_random_chunk_bitmap` was added to these layers:
- buffer
- multi buffer
- custom_highlights
- inlay_map
- fold_map
## Benchmarking
I setup benchmarks with criterion that is runnable via `cargo bench -p
editor --profile=release-fast`. When benchmarking I had my laptop
plugged in and did so from the terminal with a minimal amount of
processes running. I'm also on a m4 max
### Results
#### To Tab Point
Went from completing 6.8M iterations in 5s with an average time of
`736.13 ns` to `683.38 ns` which is a `-7.1875%` improvement
#### To Fold Point
Went from completing 6.8M iterations in 5s with an average time of
`736.55 ns` to `682.40 ns` which is a `-7.1659%` improvement
#### Editor render
Went from having an average render time of `62.561 µs` to `57.216 µs`
which is a `-8.8248%` improvement
#### Build Buffer with one long line
Went from having an average buffer build time of `3.2549 ms` to `3.2635
ms` which is a `+0.2151%` regression within the margin of error
#### Editor with 1000 multi cursor input
Went from having an average edit time of `133.05 ms` to `122.96 ms`
which is a `-7.5776%` improvement
Release Notes:
- N/A
---------
Co-authored-by: Remco Smits <djsmits12@gmail.com>
Co-authored-by: Cole Miller <cole@zed.dev>
Co-authored-by: Piotr Osiewicz <24362066+osiewicz@users.noreply.github.com>
This PR fixes the backwards compatibility of the new `Plan` variants.
We can't add new variants to the wire representation, as old clients
won't be able to understand them.
Release Notes:
- N/A
Closes#19816
Release Notes:
- Improved `ctrl-k` (`editor::CutToEndOfLine`) behavior when used at the
end of lines
- Add option to make `editor::CutToEndOfLine` not gobble newlines.
```json
{
"context": "Editor",
"bindings": { "ctrl-k": ["editor::CutToEndOfLine", { "stop_at_newlines":
true }] }
},
```
---------
Co-authored-by: Peter Tripp <peter@zed.dev>
This setting controls which visible characters are used to render
whitespace when the show_whitespace setting is enabled.
Release Notes:
- Added `whitespace_map` setting to control which visible characters are
used to render whitespace when the `show_whitespace` setting is enabled.
---------
Co-authored-by: Nia Espera <nia@zed.dev>
Whilst looking into adding support for RainbowBrackes, we stumbled upon
this: Whereas for all properties during this blending, we take the value
of `other` if it is set, for the color we actually take `self.color`
instead of `other.color` if `self.color` is at full opacity.
`Hsla::blend` returns the latter color if it is at full opacity, which
seems wrong for this case. Hence, this PR swaps these.
Will not merge before the next release, to ensure that we don't break
something somewhere unexpected.
Release Notes:
- N/A
Previously we would still show a completion menu even when the user
typed an unrecognised mode with an argument,
e.g. `@something word`.
This PR ensures that we only show the completion menu, when the part
after the `@` is a known mode (e.g. `file`/`symbol`/`rule`/...)
Release Notes:
- Fix an issue where completions for `@mentions` in the agent panel
would sometimes not be dismissed when typing a space
TODO:
- [x] Double-check if we like the naming of the new actions
- [x] Only show keybinding hint once per option (e.g. if there are two
`allow_once` buttons only show it on the first one)
- [x] If there are multiple tool calls that need authorisation, only
show keybindings on the first tool call
- [x] Figure out which keybindings to use
- [x] Add linux keybindings
- [x] Add windows keybindings
- [x] Bug: long keybindings can make the buttons overflow
Release Notes:
- Add keybindings for authorizing tool calls (`agent: Allow once`,
`agent: Allow always`, `agent: Reject once`) in the agent panel
---------
Co-authored-by: Agus Zubiaga <agus@zed.dev>
Release Notes:
- Fixed matching bracket highlighting not highlighting closing brackets
when adjacent to each other
Co-authored-by: Finn Evers <finn@zed.dev>
Potentially fixes#37367. Just going ahead with the change even though
it's unclear whether this is the fix, since it is quite low risk.
Release Notes:
- N/A
Adding mistral nemo, sorting by name and adding comment about clamp
sizing
Release Notes:
- Added support for mistral-nemo
---------
Co-authored-by: Conrad Irwin <conrad.irwin@gmail.com>
Skipping Linux and Windows keymaps here, as it's hard to find a
consistent base-binding for all 4 actions across all platforms that
don't break important actions for each keymap. Someone else can think on
that and make a proposal.
Release Notes:
- Added bindings for navigating between sibling nodes in the syntax tree
on macOS (`cmd-ctrl-{up/down}` for `SelectPreviousSyntaxNode` and
`SelectNextSyntaxNode`). Breaking change: the existing syntax tree
parent/child navigation bindings have moved from
`ctrl-shift-{left/right}` to `cmd-ctrl-{left/right}` to create a unified
four-directional navigation pattern where all syntax tree operations use
the same modifier combination. We could not use the previous base
modifiers without breaking more bindings.
Closes#37208
## Release Notes:
- Fixed: SVG preview now refreshes automatically when files are modified
by external programs
## Summary
Previously, SVG preview would only refresh when files were saved within
the Zed editor, but not when modified by external programs (like
scripts, other editors, etc.)
## What Changed
The SVG preview now subscribes to file system events through the
worktree system. When an external program modifies an SVG file, the
worktree detects the change and notifies the preview. The preview then
clears its cache and refreshes to show the updated content.
## Before the fix
https://github.com/user-attachments/assets/e7f9a2b2-50f9-4b43-95e9-93a0720749f5
## After the fix
https://github.com/user-attachments/assets/b23511e3-8e59-45a1-b29b-d5105d32bd2c
AI Usage:
Used Cursor for code generation
While working on fixing this: #37116. I reliased the current
implementation of github copilot is not truly resilient to upstream
changes.
This PR enhances GitHub Copilot Chat to be forward-compatible with new
AI model vendors and improves token counting accuracy by using
vendor-specific tokenizers from the GitHub Copilot API. The system
previously failed when GitHub added new model vendors like xAI with
deserialization errors, and token counting wasn't utilizing the
vendor-specific tokenizer information provided by the API. The solution
adds an Unknown variant to the ModelVendor enum with serde other
attribute to gracefully handle any new vendors GitHub introduces,
implements tokenizer-aware token counting that uses the model's
specified tokenizer mapping o200k_base to gpt-4o with fallback, adds
explicit support for xAI models with proper tool input format handling,
and includes comprehensive test coverage for unknown vendor scenarios.
Key changes include adding the tokenizer field to model capabilities,
implementing the tokenizer method on models, updating tool input format
logic to handle unknown vendors, and simplifying token counting to use
the vendor's specified tokenizer or fall back to gpt-4o. This ensures
Zed's Copilot Chat integration remains robust and accurate as GitHub
continues expanding their AI model provider ecosystem.
Release Notes:
- Enhanced model vendor compatibility to automatically support future AI
providers and improved token counting accuracy using vendor-specific
tokenizers from the GitHub Copilot
---------
Signed-off-by: Umesh Yadav <git@umesh.dev>
Closes#37511
The outline modal seems to have a bug where if it's open and the
`outline::Toggle` is triggered, it would not close if there was another
command with the same keybind. So instead, if the outline modal is open
and an `outline::Toggle` is triggered, we dismiss the modal.
Release Notes:
- Fixed a bug where `outline::Toggle` would sometimes not close outline
modal
Somehow we have a regression where the auth wasn't being called, so the
model didn't exist.
Looking at the code, it is likely this was relying on some other part of
the code doing the auth, since the order wouldn't have worked before
without that happening. This new order of doing auth before checking for
available models should fix it going forward.
Release Notes:
- N/A
The "Install CLI" menu button and `install_cli::Install` action are
effectively no-op on Windows since the CLI is directly available in
Windows Terminal (CMD prompt, PowerShell, etc.) after the user runs Zed
installer package.
Release Notes:
- N/A
Co-authored-by: Peter Tripp <peter@zed.dev>
This is an initial implementation that isn't used for any settings yet,
but will be used once `Vec<String>` is implemented.
I also updated the window.with_state api to grant access to a
`Context<S>` app reference instead of just an App.
## Example
<img width="603" height="83" alt="Screenshot 2025-09-09 at 2 15 56 PM"
src="https://github.com/user-attachments/assets/7b3fc350-a157-431f-a4bc-80a1806a3147"
/>
Release Notes:
- N/A
Closes#21193Closes#14703
Having the ability to navigate directly to the next
symbolHighlight/reference lets you follow the data flow of a variable.
If you highlight the function itself (depending on the LSP), you can
also navigate to all returns.
Note that this is a different feature from navigating to the next match,
as that is not language-context aware. For example, if you have a var
named foo it would also navigate to an unrelated variable fooBar.
Here's how this patch works:
- The editor struct has a background_highlights.
- Collect all highlights with the keys [DocumentHighlightRead,
DocumentHighlightWrite]
- Depending on the direction, move the cursor to the next or previous
highlight relative to the current position.
Release Notes:
- Added `editor::GoToNextDocumentHighlight` and
`editor::GoToPreviousDocumentHighlight` to navigate to the next LSP
document highlight. Useful for navigating to the next usage of a certain
symbol.
When the cursor was sitting on a syntactically insignificant character,
like a `{` or `,`, this function was selecting only that character, when
what the user likely wanted was to select the next larger syntax node.
Those punctuation characters all seemed to be not "named", in
tree-sitter terminology, so I updated the function to walk up the node
tree until it found a node where `is_named()` is true.
Closes#4555
Also, while writing the tests, the output of a failing test with the
wrong thing selected was harder to read than it needed to be.
It used to output a diff of ranges, like this:
<img width="217" height="111" alt="image"
src="https://github.com/user-attachments/assets/00de53a8-8776-47aa-8101-5a5b5bc3fa5e"
/>
I leveraged the existing `generate_marked_text` helper function and
updated the assertion to output a diff of the text with the selection
markers:
<img width="211" height="116" alt="image"
src="https://github.com/user-attachments/assets/53b2b882-2676-4c70-8718-e2e2ba6f254e"
/>
Happy to make that a separate PR, if needed.
Release Notes:
- Fixed Editor select_larger_syntax_node to be smart about punctuation.
Adds a new menu option to toggle between sorting git entries by path or
status, with settings integration
Release Notes:
- Git Panel: Added toggle to switch between sorting git panel entries by
path or by status (available in git panel ellipsis menu)
---------
Co-authored-by: Conrad Irwin <conrad.irwin@gmail.com>
It appears that in macOS, the `AcpConnection._wait_task` doesn't always
get dropped when quitting the app. In these cases, the subprocess would
be kept alive because we move the `child` into it.
Instead, we will now explicitly kill it when `AcpConnection` is dropped.
It's ok to do this because when the connection is dropped, the thread is
also dropped, so there's no need to report the exit status to it.
Closes#37741
Release Notes:
- Claude Code: Fix subprocess leak on app quit
Follow up to this: #35114
* Previously we were still showing the commit message even after
cancelling amend which was the old commit message.
* This PR fixes that by restoring the commit message to the old state
before the amend begin so that in case user typed a commit message it's
shown if not then it's not.
Before:
https://github.com/user-attachments/assets/e0edcfff-863d-4367-a7c4-8a2998e702ca
After:
https://github.com/user-attachments/assets/9daf9be0-8a3d-4914-91a4-567693711b6b
Release Notes:
- Resolved an issue where cancelling an amend commit would incorrectly
leave the old commit message in the input field. The commit message box
now properly reverts to its pre-amend state.
David Kleingeld