### DISCLAIMER
> As of 6th March 2025, debugger is still in development. We plan to
merge it behind a staff-only feature flag for staff use only, followed
by non-public release and then finally a public one (akin to how Git
panel release was handled). This is done to ensure the best experience
when it gets released.
### END OF DISCLAIMER
**The current state of the debugger implementation:**
https://github.com/user-attachments/assets/c4deff07-80dd-4dc6-ad2e-0c252a478fe9https://github.com/user-attachments/assets/e1ed2345-b750-4bb6-9c97-50961b76904f
----
All the todo's are in the following channel, so it's easier to work on
this together:
https://zed.dev/channel/zed-debugger-11370
If you are on Linux, you can use the following command to join the
channel:
```cli
zed https://zed.dev/channel/zed-debugger-11370
```
## Current Features
- Collab
- Breakpoints
- Sync when you (re)join a project
- Sync when you add/remove a breakpoint
- Sync active debug line
- Stack frames
- Click on stack frame
- View variables that belong to the stack frame
- Visit the source file
- Restart stack frame (if adapter supports this)
- Variables
- Loaded sources
- Modules
- Controls
- Continue
- Step back
- Stepping granularity (configurable)
- Step into
- Stepping granularity (configurable)
- Step over
- Stepping granularity (configurable)
- Step out
- Stepping granularity (configurable)
- Debug console
- Breakpoints
- Log breakpoints
- line breakpoints
- Persistent between zed sessions (configurable)
- Multi buffer support
- Toggle disable/enable all breakpoints
- Stack frames
- Click on stack frame
- View variables that belong to the stack frame
- Visit the source file
- Show collapsed stack frames
- Restart stack frame (if adapter supports this)
- Loaded sources
- View all used loaded sources if supported by adapter.
- Modules
- View all used modules (if adapter supports this)
- Variables
- Copy value
- Copy name
- Copy memory reference
- Set value (if adapter supports this)
- keyboard navigation
- Debug Console
- See logs
- View output that was sent from debug adapter
- Output grouping
- Evaluate code
- Updates the variable list
- Auto completion
- If not supported by adapter, we will show auto-completion for existing
variables
- Debug Terminal
- Run custom commands and change env values right inside your Zed
terminal
- Attach to process (if adapter supports this)
- Process picker
- Controls
- Continue
- Step back
- Stepping granularity (configurable)
- Step into
- Stepping granularity (configurable)
- Step over
- Stepping granularity (configurable)
- Step out
- Stepping granularity (configurable)
- Disconnect
- Restart
- Stop
- Warning when a debug session exited without hitting any breakpoint
- Debug view to see Adapter/RPC log messages
- Testing
- Fake debug adapter
- Fake requests & events
---
Release Notes:
- N/A
---------
Co-authored-by: Piotr Osiewicz <24362066+osiewicz@users.noreply.github.com>
Co-authored-by: Anthony Eid <hello@anthonyeid.me>
Co-authored-by: Anthony <anthony@zed.dev>
Co-authored-by: Piotr Osiewicz <peterosiewicz@gmail.com>
Co-authored-by: Piotr <piotr@zed.dev>
I vibe coded this in Zed, dawg.
This avoids a number of cases where we created multiple copies of the
same lsp adapter. Now we clone the Arcs.
Release Notes:
- N/A
Similar to how tasks are fetched via LSP, also queries for document's
code lens and filters the ones with the commands, supported in server
capabilities.
Whatever's left and applicable to the range given, is added to the
actions menu:
This way, Zed can get more actions to run, albeit neither r-a nor vtsls
seem to provide anything by default.
Currently, there are no plans to render code lens the way as in VSCode,
it's just the extra actions that are show in the menu.
------------------
As part of the attempts to use rust-analyzer LSP data about the
runnables, I've explored a way to get this data via standard LSP.
When particular experimental client capabilities are enabled (similar to
how clangd does this now), r-a starts to send back code lens with the
data needed to run a cargo command:
```
{"jsonrpc":"2.0","id":48,"result":{"range":{"start":{"line":0,"character":0},"end":{"line":98,"character":0}},"command":{"title":"▶︎ Run Tests","command":"rust-analyzer.runSingle","arguments":[{"label":"test-mod tests::ecparser","location":{"targetUri":"file:///Users/someonetoignore/work/ec4rs/src/tests/ecparser.rs","targetRange":{"start":{"line":0,"character":0},"end":{"line":98,"character":0}},"targetSelectionRange":{"start":{"line":0,"character":0},"end":{"line":98,"character":0}}},"kind":"cargo","args":{"environment":{"RUSTC_TOOLCHAIN":"/Users/someonetoignore/.rustup/toolchains/1.85-aarch64-apple-darwin"},"cwd":"/Users/someonetoignore/work/ec4rs","overrideCargo":null,"workspaceRoot":"/Users/someonetoignore/work/ec4rs","cargoArgs":["test","--package","ec4rs","--lib"],"executableArgs":["tests::ecparser","--show-output"]}}]}}}
```
This data is passed as is to VSCode task processor, registered in
60cd01864a/editors/code/src/main.ts (L195)
where it gets eventually executed as a VSCode's task, all handled by the
r-a's extension code.
rust-analyzer does not declare server capabilities for such tasks, and
has no `workspace/executeCommand` handle, and Zed needs an interactive
terminal output during the test runs, so we cannot ask rust-analyzer
more than these descriptions.
Given that Zed needs experimental capabilities set to get these lens:
60cd01864a/editors/code/src/client.ts (L318-L327)
and that the lens may contain other odd tasks (e.g. docs opening or
references lookup), a protocol extension to get runnables looks more
preferred than lens:
https://rust-analyzer.github.io/book/contributing/lsp-extensions.html#runnables
This PR does not include any work on this direction, limiting to the
general code lens support.
As a proof of concept, it's possible to get the lens and even attempt to
run it, to no avail:
Release Notes:
- Used `textDocument/codeLens` data in the actions menu when applicable
Partially addresses #25698
Part of why autocomplete suggestions for `keymap.json` aren't great is
because `:` is (correctly) considered a punctuation character, rather
than a word character, in JSON. But since `::` is part of the name of
zed commands, it means that the autocomplete context window loses
context after the user types colon:
Suggestion here is to use overrides for JSON and JSONC such that colon
is considered a word character when it's inside a string. This improves
the experience:
I believe this is more broadly correct anyway, since `:` loses it's
punctuation meaning when inside a string.
Hope this is helpful!
Release Notes:
- Improved autocomplete for keymap.json by treating `::` like word characters when inside a string.
Because of #26562, it is now possible to subscribe to extension update
events within the LSP store, where we can then update the Schemas sent
to the JSON LSP resulting in dynamic updates to the auto-complete
suggestions and diagnostics in settings. Notably, this means newly
installed languages and (icon) themes will auto-complete correctly as
soon as the extension is installed.
Closes#15436
Release Notes:
- Fixed an issue where autocomplete suggestions and diagnostics for
languages and (icon) themes in settings would not update when the
extension with which they were added was installed or uninstalled
Disables https://github.com/zed-industries/zed/pull/26146 until a better
way to add diagnostics is found.
Overall, the PR had made changes that are worth keeping instead of
reverting, such as finally extracting out r-a's language server logic
into an `_ext.rs` file.
Release Notes:
- N/A
Reverts zed-industries/zed#25135
This approach was not the best as explained in the response to the
original PR. Likely, the better approach is to create a newer specific
scope for these kinds of variables under the `@variable` prefix so that
themes can control these pseudo-keywords specifically
Closes#4271
Implemented by kicking of a task on the main thread at the end of
`Editor::handle_input` which waits for the buffer to be re-parsed before
checking if JSX tag completion possible based on the recent edits, and
if it is then it spawns a task on the background thread to generate the
edits to be auto-applied to the buffer
Release Notes:
- Added support for auto-closing of JSX tags
---------
Co-authored-by: Cole Miller <cole@zed.dev>
Co-authored-by: Max Brunsfeld <max@zed.dev>
Co-authored-by: Marshall Bowers <git@maxdeviant.com>
Co-authored-by: Mikayla <mikayla@zed.dev>
Co-authored-by: Peter Tripp <peter@zed.dev>
Fixes#16057
In newer versions of clangd, the switch labelDetailsSupport in the json
passed to the language server modifies the format of the returned json.
Zed handles well the old format, but misses the function parameters in
the new one. For example:
The old format looks like this:
```json
...
"label": " Window(int width, int height, const char *name, bool vsync, bool resizable)",
...
```
and with labelDetailsSupport = true:
```json
...
"label": " Window",
"labelDetails": {
"detail": "(int width, int height, const char *name, bool vsync, bool resizable)"
},
...
```
A simple solution is to just to not tell the language server that label
details are supported and force it to use the old format. This is a
dirty fix, but makes the completions behave like in the old versions of
clangd.
I do not know if this will break another language server. From what I've
found out most lsp-s do not depend on that setting and provide all
completion data either way. If not, this switch will need to be exposed
in a config or be at least lsp-dependant.
Lastly, I do not know Rust, maybe will need help to make a better fix
for the issue.
Release Notes:
- Fixed broken C++ completion suggestions
In #25005 we added regex syntax highlighting to search; but the existing
regex grammar highlighted every character as a string which was hard to
read.
This flips so that characters are not highlighted, and brackets, etc.
are.
<img width="346" alt="Screenshot 2025-03-03 at 14 39 35"
src="https://github.com/user-attachments/assets/f7d3ae9c-fb5c-45eb-a5e9-41a330fbe940"
/>
Release Notes:
- Fixed regex search box being overly green
Closes#19267
Adds highlight field specifically for `constructor` in JS/TS/TSX so that
it can be highlighted as a different color than regular methods
Release Notes:
- N/A
---------
Co-authored-by: Ben Kunkle <ben@zed.dev>
#25333 added broader highlighting for identifiers, which broke the
generic query for attribute queries, resulting in these being
highlighted the same as identifiers.
To accomodate for this change, this PR updates the attribute matches to
be more specific.
Additionally, path matches in scoped identifiers are no longer
highlighted as attributes, as seen in the comparison screenshot. Can
revert this if requested.
| Zed Preview | <img width="750" alt="preview"
src="https://github.com/user-attachments/assets/2cd2e830-f510-4adf-8ce9-c41ed6fb157c"
/> |
| --- | --- |
| `main` | <img width="750" alt="main"
src="https://github.com/user-attachments/assets/cbe93186-9afd-4515-bc06-e519fd4ee6af"
/> |
| This PR | <img width="750" alt="pr"
src="https://github.com/user-attachments/assets/68270de8-e083-4fc6-a45e-25d3151acd87"
/> |
The generic match for `token_tree` is needed to recursively match
patterns like `#[cfg(any(test, feature = "test-support"))]` (or at least
I was unable to find a better query here). I tried to validate that this
does not break any other highlights and I believe it does not. However,
I might have still missed something.
Release Notes:
- N/A
This PR is an attempt to add support for `--target-dir` argument to
`cargo` commands when executing tasks with rust.
When using VSCode I was already using this trick to not block the
current binary compilation when I was trying a specific test. As it's a
different target directory it won't block the `cargo` commands I'm using
in my terminal.
I used the task variables to achieve this but I'm not sure it's the best
option to be honest. I didn't find any examples in your docs to see if
sometimes you had specific configuration for languages and tasks.
Let me know if this solution would be a good fit and if the
implementation is ok.
If so feel free to redirect me to an example I can reproduce to write a
unit test or so... And I will also update the docs.
Example of config:
```
{
"languages": {
"Rust": {
"tasks": {
"variables": {
"RUST_TARGET_DIR": ".cargo_check"
}
}
}
}
}
```
it will run `cargo test -p XXX --target-dir .cargo-check`
Release Notes:
- Added support for `--target-dir` for Rust tasks
---------
Signed-off-by: Benjamin <5719034+bnjjj@users.noreply.github.com>
Release Notes:
- Fixed issue where `true` and `false` were highlighted as constants,
ignoring the `boolean` highlight defined in themes.
- This fix applies to: C, C++, Go, JSON, JSONC, Python, and Rust.
---------
Co-authored-by: Marshall Bowers <git@maxdeviant.com>
Screenshot:
I would be happy to add tests if you point me to the right place to do
it please.
Release Notes:
- Added support for doc test in tasks for Rust
---------
Signed-off-by: Benjamin <5719034+bnjjj@users.noreply.github.com>
Closes#24565
As pointed out in
https://github.com/zed-industries/zed/issues/24565#issuecomment-2657822723
, the name for the Pyright language server is `pyright-langserver`, not
`pyright`. The latter is a CLI-tool as described in
https://microsoft.github.io/pyright/#/command-line which only provides
static type checking. It has neither LSP-capabilities nor a `--stdio`
argument. Thus, the error as shown in the linked issue appears.
I disagree with the fix as described in
https://github.com/zed-industries/zed/issues/24565#issuecomment-2657904208
, as it could only cause this error to reappear in rare scenarios where
Pyright, but not the Pyright language server is installed in a user's
environment. Just checking for `pyright-langserver` to be present in the
environment seems more straightforward here.
Release Notes:
- Python: Fixed Pyright failing to start when installed locally
Co-authored-by: Beniamin Zagan <47153906+beniaminzagan@users.noreply.github.com>
Release Notes:
- Added support for checking for `package-version-server` on the
`$PATH`.
---------
Signed-off-by: Matthew Penner <me@matthewp.io>
Co-authored-by: Marshall Bowers <git@maxdeviant.com>
Closes#24951
We were highlighting both as `@variable.special` however, they are
_techinically_ keywords and other editors (VSCode/WebStorm) seem to
highlight them as keywords as well.
Release Notes:
- N/A
### Overview
This PR improves the existing
[mini.ai‐like](https://github.com/echasnovski/mini.ai) text-object logic
for both “AnyQuotes” (quotes) and “AnyBrackets” (brackets) by adding a
multi‐line fallback. The first pass searches only the current line for a
best match (cover or next); if none are found, we do a multi‐line pass.
This preserves mini.ai's usual “line priority” while ensuring we can
detect pairs that start on one line and end on another.
### What Changed
1. Brackets
- Line-based pass uses `gather_line_brackets(map, caret.row()) `to find
bracket pairs `((), [], {}, <>) `on the caret’s line.
- If that fails, we call `gather_brackets_multiline(map)` to single‐pass
scan the entire buffer, collecting bracket pairs that might span
multiple lines.
- Finally, we apply the mini.ai “**cover or next**” logic
(`pick_best_range`) to choose the best.
2. Quotes
- Similar line-based pass with `gather_line_quotes(map, caret.row())`.
- If no local quotes found, we do a multi‐line fallback with
`gather_quotes_multiline(map)`, building a big string for the whole
buffer and using naive regex for "...", '...', and `...`.
- Also preserves “inner vs. outer” logic:
- For inner (e.g. `ciq`), we skip bounding quotes or brackets if the
range is at least 2 characters wide.
- For outer (`caq`), we return the entire range.
3. Shared “`finalize`” helpers
- `finalize_bracket_range` and `finalize_quote_range` handle the “inner”
skip‐chars vs. “outer” logic.
- Both rely on the same “line first, then full fallback” approach.
### Why This Matters
- **Old Behavior**: If you had multi‐line brackets { ... } or multi‐line
quotes spanning multiple lines, they weren’t found at all, since we only
scanned line by line. That made text objects like ci{ or ciq fail in
multi-line scenarios.
- **New Behavior**: We still do a quick line pass (for user‐friendly
“line priority”), but now if that fails, we do a single‐pass approach
across the entire buffer. This detects multi‐line pairs and maintains
mini.ai’s “cover‐or‐next” picking logic.
### Example Use Cases
- **Curly braces:** e.g., opening { on line 10, closing } on line 15 →
previously missed; now recognized.
- **Multi‐line quotes**: e.g., "'Line 1\nLine 2', no longer missed. We
do gather_quotes_multiline with a naive regex matching across newlines.
### Tests
- Updated and expanded coverage in:
- test_anyquotes_object:
- Includes a multi-line '...' test case.
- E.g. 'first' false\n<caret>string 'second' → ensuring we detect
multi‐line quotes.
- test_anybrackets_object:
- Verifies line‐based priority but also multi‐line bracket detection.
- E.g., an open bracket ( on line 3, close ) on line 5, which used to
fail.
### Limitations / Future Enhancements
- **Escaping**: The current approach for quotes is naive and doesn’t
handle escape sequences (like \") or advanced parser logic. For deeper
correctness, we’ll need more advanced logic, this is also not supported
in the original mini.ai plugin so it is a known issue that won't be
attended for now.
### Important Notes
- Fix for the bug: https://github.com/zed-industries/zed/issues/23889
this PR addresses that bug specifically for the AnyQuotes text object.
Note that the issue still remains in the built-in motions (ci', ci",
ci`).
- Caret Position Differences: The caret position now slightly deviates
from Vim’s default behavior. This is intentional. I aim to closely mimic
the mini.ai plugin. Because these text objects are optional
(configurable via vim.json), this adjusted behavior is considered
acceptable and in my opinion the new behavior is better and it should be
the default in vim. Please review the new tests for details and context.
- Improved Special Cases: I’ve also refined how “false strings” in the
middle and certain curly-bracket scenarios are handled. The test suite
reflects these improvements, resulting in a more seamless coding
experience overall.
### References:
- Mini.AI plugin in nvim: https://github.com/echasnovski/mini.ai
Thank you for reviewing these changes!
Release Notes:
- Improve logic of aq, iq, ab and ib motions to work more like mini.ai
plugin
Closes#14497
Release Notes:
- Fixed outline view for Go to correctly indent types and show missing
variables.
----
This PR fixes the tree-sitter query for outlines for Go code. It
correctly indents "grouped" `type` declarations. It also fixes missing
variables for "grouped" `var` declarations.
### Before:
1. Incorrectly indented types from `G` to `K` and `aliasInt`
2. Missing vars `M` and `N` in outline
### After:
1. Types no longer indented incorrectly
2. Missing vars `M` and `N` appear in the outline, as expected
### Caveats:
1. This fix comes from an hour or so of reading about tree-sitter and
its query syntax. I'm not an expert.
2. I'm not sure how to test this. I've done manual testing and it
appears to works as expected without an regressions.
