This PR updates the `push_div` method in the `MarkdownElementBuilder` to
support taking in a `Stateful<Div>`.
This is some groundwork for supporting horizontal scrolling in Markdown
code blocks.
Release Notes:
- N/A
Closes#19837
This PR is a continuation of [linux: Implement
Menus](https://github.com/zed-industries/zed/pull/21873) and should only
be reviewed once the existing PR is merged.
I created this as a separate PR as the existing PR was already reviewed
but is yet to merge, and also it was my initial plan to do it in
separate parts because of the scope of it. This will also help reviewing
code faster.
This PR adds two new types of keyboard shortcuts to make menu navigation
easier:
1. `Alt + Z` for Zed, `Alt + F` for File, `Alt + S` for Selection, and
so on to open a specific menu with this combination. This mimics VSCode
and IntelliJ.
2. `Arrow Left/Right` when any menu is open. This will trigger the
current menu to close, and the previous/next to open respectively. First
and last element cycling is handled.
`Arrow Up/Down` to navigate menu entries is already there in existing
work.
https://github.com/user-attachments/assets/976aea48-4e20-4c19-850d-4d205a4bead2
Release Notes:
- Added keyboard navigation for menus on Linux (left/right). If you wish
to open menus with keyboard shortcuts add the following to your user
keymap:
```json
{
"context": "Workspace",
"bindings": {
"alt-z": ["app_menu::OpenApplicationMenu", "Zed"],
"alt-f": ["app_menu::OpenApplicationMenu", "File"],
"alt-e": ["app_menu::OpenApplicationMenu", "Edit"],
"alt-s": ["app_menu::OpenApplicationMenu", "Selection"],
"alt-v": ["app_menu::OpenApplicationMenu", "View"],
"alt-g": ["app_menu::OpenApplicationMenu", "Go"],
"alt-w": ["app_menu::OpenApplicationMenu", "Window"],
"alt-h": ["app_menu::OpenApplicationMenu", "Help"]
}
}
```
---------
Co-authored-by: Peter Tripp <peter@zed.dev>
This PR adds another example thread to showcase a response with long
lines of code.
This example will be helpful when working to make the code blocks scroll
horizontally instead of wrapping.
Release Notes:
- N/A
Clicking buffer headers and line numbers would sometimes take you to a
disorienting scroll position. This PR improves that so the destination
line is roughly at the same Y position as it appeared in the
multibuffer.
https://github.com/user-attachments/assets/3ad71537-cf26-4136-948f-c5a96df57178
**Note**: The alignment won't always be perfect because the multibuffer
and target buffer might start at a different absolute Y position
(because of open search, breadcrumbs, etc). I wanted to compensate for
that, but that requires a fundamental change that I'd prefer to make
separately.
Release Notes:
- Fix vertical alignment when jumping from multibuffers
This PR is a follow-up to #22200 that makes it so the chat panel icon is
visible when the chat panel is active, even if not in a call (when using
the `when_in_call` setting).
Release Notes:
- N/A
* Skip walking string for truncate when byte len is <= char limit
* Fix `truncate_and_remove_front` returning string that is `max_chars +
1` in length. Now more consistent with `truncate_and_trailoff` behavior.
* Fix `truncate_and_remove_front` adding ellipsis when max_chars == char
length
Release Notes:
- N/A
Closes#19974.
When a file is pasted in the project panel at a location where a file
with that name already exists, the new file's name is disambiguated by
appending " copy" at the end. This happens on the paste and the
duplicate actions, as well as when Alt-dragging files.
With this PR, this will now open the file rename editor with the
disambiguator pre-selected.
Open question:
With this PR's current implementation, this won't always work when
pasting multiple files at once. In this case, the file rename editor
only opens for the last pasted file, if that file was disambiguated. If
only other files were disambiguated instead, it won't open.
This roughly mimics the previous paste behaviour, namely that only the
last pasted file was selected.
I see two options here: If multiple files were pasted and some of them
were disambiguated, we could select and open the rename editor for the
last file that was actually disambiguated (easy), or we could open a
kind of multi-editor for all files (hard, but maybe a multi-rename
editor could actually be interesting in general...).
Release Notes:
- Open rename file editor if pasted file was disambiguated
Closes#10325
Release Notes:
- Fixed an issue inside the integrated terminal where clicking on URLs
that started with `file://` would sometimes not work when the path
included a line number (e.g. `file:///Users/someuser/lorem.txt:221:22`)
This PR updates the scrollbar diagnostic setting to provide fine-grained
control over which indicators to show, based on severity level. This
allows the user to hide lower-severity diagnostics that can otherwise
clutter the scrollbar (for example, unused or disabled code).
The options are set such that the existing boolean setting has the same
effect: when `true` all diagnostics are shown, and when `false` no
diagnostics are shown.
Closes#22296.
Release Notes:
- Added fine-grained control of scrollbar diagnostic indicators.
This commit is all about strings, not about line layout at all. When
laying out text, we use a line layout cache to avoid roundtrips to
system layout engine where possible. This makes it so that we might end
up not needing an owned version of text to insert into the cache, as we
might get a cached version.
The API boundary of line layout accepted text to be laid out as &str. It
then performed cache lookup (which didn't require having an owned
version) and only resorted to making an owned version when needed. As it
turned out though, exact cache hits are quite rare and we end up needing
owned version more often than not. The callers of line layout either
dealt with SharedStrings or owned Strings. Due to coercing them into
&str, we were ~always copying text into a new string (unless there was a
same-frame-hit). This is a bit wasteful, thus this PR generifies the API
a bit to make it easier to reuse existing string allocations if there
are any.
Benchmark scenario: scrolling down page-by-page through editor_tests (I
ran the same scenario twice):
Release Notes:
- N/A
Closes https://github.com/zed-industries/zed/issues/21657
Follow-up of https://github.com/zed-industries/zed/pull/22488
Previous PR broke git blame tooltips, which are expected to be open when
hovered, even if the mouse cursor is moved away from the actual blame
entry that caused the tooltip to appear.
Current version moves the invalidation logic into `prepaint_tooltip`,
where the new data about the tooltip origin is used to ensure we
invalidate only tooltips that have no mouse cursor in either origin
bounds or tooltip bounds (if it's hoverable).
Release Notes:
- Fixed tooltips getting stuck
This reverts commit 344284e013.
That change broke git blame tooltips, as Zed should also show tooltips
which are hovered, even though the mouse had left the origin element's
bounds.
Release Notes:
- N/A
This pull request includes a small change to the
`crates/languages/src/markdown/config.toml` file. The change adds block
comment syntax for Markdown files.
*
[`crates/languages/src/markdown/config.toml`](diffhunk://#diff-4cf73d9af0f11f2ac8929bd8113ee76aa382dc96a731f18510c09fc3d0db1f9cR5):
Added block comment syntax `<!-- ` and ` -->` for Markdown files.
Closes https://github.com/zed-industries/zed/issues/19797
I have been testing it with the following:
1. create a simple markdown file containing following words
```markdown
good morning - English
Dzień dobry - polish
おはよう - japanese
صباح الخير - arabic
早安 - Chineses (Traditional)
```
2. comment using `cmd+/` and see if there are any errors
Release Notes:
- Added block comment syntax `<!-- ` and ` -->` for Markdown files.
Closes https://github.com/zed-industries/zed/issues/22469 — and, aside
from fixing that issue also took the opportunity to make the
`file_search` icon dimensions feel better in comparison the other icons
in the Project Search area.
Release Notes:
- N/A
For both the strings and paths multithreaded matching it would still
aggregate the response even though it is unneeded. It now checks cancel.
In the paths matcher, cancel is now checked within the loop, since it
was calling `match_candidates` even though no further results would be
computed.
Release Notes:
- N/A
* Removes `max_results` from the matcher interface as this is better
dealt with in consumers once all results are known. The current
implementation was quite inefficient as it was using binary search to
find insertion points and then doing an insert which copies the entire
suffix each time.
* There was a corner case where if the binary search found a match
candidate with the same score, it was dropped. Now fixed.
* Uses of `util::extend_sorted` when merging results from worker threads
also repeatedly uses binary search and insertion which copies the entire
suffix. A followup will remove that and its usage.
* Adds `util::truncate_to_bottom_n_sorted_by` which uses quickselect +
sort to efficiently get a sorted count limited result.
* Improves interface of Matcher::match_candidates by providing the match
positions to the build function. This allows for removal of the `Match`
trait. It also fixes a bug where the Match's own Ord wasn't being used,
which seems relevant to PathMatch for cases where scores are the same.
Release Notes:
- N/A
Part of https://github.com/zed-industries/zed/issues/22415
I've noticed that I cannot work with any Zig projects, as there were no
"go to definition", formatting and inlay hints.
After debugging, I've discovered that `typos` was registered as a first
language server, becoming the "primary" one for Zig.
That one does not have any proper capabilities, hence all corresponding
LSP requests were no-op.
While this solution is not ideal (I wonder, how many other set-ups are
broken due to the same thing?), we'd better fix things for now this way
at least.
Release Notes:
- Fixed `zls` not working properly when `typos` extension is installed
When switching from just running the editor directly to using the CLI I
missed the logs in the terminal and started using `--foreground`. I
realized this was because there was a tiny amount of effort involved in
finding out where logs were being written to. Having the cli output it
to stderr helps make this more visible.
Seems like such a minor thing not listing in release notes.
Release Notes:
- N/A
This means that `workspace::ToggleRightDock` will open the assistant if
no right-dock panel has been manually activated, instead of the chat as
before. Also cleans up the `active_panel_index` logic a bit.
cc @nathansobo
Release Notes:
- Make `workspace::ToggleRightDock` open the assistant panel if no
right-dock panel has yet been activated
Changes:
- [x] Increase expand affordance surface area
- [x] Ensure expand buttons have tooltips with keybindings
- [x] Make line numbers clickable to jump you to location (only in
multibuffers)
- [x] Hide the "Jump To File" element in not-focused excerpts
Before merging it:
- [x] Fix off-by-one header focus styles glitch
Improvements to consider for follow-up PRs:
1. Experiment with increasing the width of the clickable surface area
for line numbers
2. Don't show (or disable) the "expand excerpt" button when at the top
or bottom edge of the file
3. Once you jump to location, centralize the cursor scroll position
Release Notes:
- Improved multibuffer's "expand excerpt" affordance
- Fixed "jump to file/location" and "expand excerpt" keybinding display
- Made clicking on line numbers in multibuffers jump you to cursor
location in file
---------
Co-authored-by: Thorsten Ball <mrnugget@gmail.com>
Co-authored-by: Agus Zubiaga <hi@aguz.me>
Co-authored-by: Kirill Bulatov <kirill@zed.dev>
Co-authored-by: Agus Zubiaga <agus@zed.dev>
This slashes our incremental dev times (touch editor) by 0.6s
(8.1->7.6s) due to unblocking terminal_view build sooner.
Closes #ISSUE
Release Notes:
- N/A
Closes https://github.com/zed-industries/zed/issues/21657
In case of the task rerun button tooltip from
f6dabadaf7/crates/terminal_view/src/terminal_view.rs (L1051-L1070)
, the actual button element is not styled as invisible, only its parent.
Zed won't render such element since it's parent is hidden, but will
consider it "visible" all the time its `paint` is called, spawning a
task with the delay, that will create the tooltip:
f6dabadaf7/crates/gpui/src/elements/div.rs (L1949-L1959)
When the parent is hidden, the child won't be painted anymore, and no
mouse listeners will be able to detect this fact and hide the tooltip.
Hence, check such cases separately, during `prepaint`, and invalidate
the tooltips that are not valid anymore.
We cannot use `hitbox.is_hovered(cx)` as it's not really hovered during
prepaint, so a mouse position check is used instead.
Release Notes:
- Fixed tooltips getting stuck
* Remove unnecessary WindowContext and ViewContext '_ lifetimes
* Removed some cases where WindowContext has a different name than `cx`.
Release Notes:
- N/A
Closes https://github.com/zed-industries/zed/issues/21906
* After https://github.com/zed-industries/zed/pull/21238,
`TerminalPanel` and `Project` strong references were moved into
`Pane`-related closures, creating a cycle, that did not allow
registering project release and shutting down corresponding language
servers
* After https://github.com/zed-industries/zed/pull/22329, a special
`Editor` was created with a strong reference to the `Project` which
seemed to do nothing bad in general, but when a working Zed was running
a Zed Dev build, had the same issue with preventing language servers
from shutting down.
The latter is very odd, and seems quite dangerous, as any arbitrary
`Editor` with `Project` in it may do the same, yet it seems that we did
not store them before the way git panel does.
I have tried creating a test, yet seems that we need to initialize a lot
of Zed for it which I failed — all my attempts resulted in a single
language server being present in the `Project`'s statuses.
Release Notes:
- Fixed language servers not being released between project reopens
Up until now, in the assistant 2, if you scrolled up either while a
message was being generated or after it's been generated, then submitted
a new message, you'd keep your scroll position. Now, with this PR, if
your scroll position is somewhere else that's not the bottom, as you
submit a new message, you'll be back at the bottom.
https://github.com/user-attachments/assets/8b111c10-27ff-4d7b-9b10-4c31093c6457
Release Notes:
- N/A
Co-authored-by: Agus Zubiaga <hi@aguz.me>
