Motivation for this change is to use `ContextStore` in headless
assistant, which requires it to not depend on UI entities like
`Workspace`.
This reapplies a change that was revert was in #28428, and fixes the panic.
Release Notes:
- N/A
PR #24388 changed the markdown parsing to copy parsed text in order to
handle markdown escaping, removing the optimization to instead reuse
text from the input.
Another issue with that change was that handling of finding links within
`Text` intermixed use of `text` and `parsed`, relying on the offsets
matching up (which I believe was true in practice).
The solution is to distinguish pulldown_cmark `Text` nodes that share
bytes with the input and those that do not.
Release Notes:
- N/A
Motivation for this is to simplify the context types and make it cleaner
to add image context.
Release Notes:
- N/A
---------
Co-authored-by: Nathan Sobo <nathan@zed.dev>
This seems more correct as corners are not necessarily only for rounding
radii.
Also applies clamping after scaling in `paint_quad`, deduplicating that
logic. This also provides a more precise result by doing the clamping
after scaling, avoiding floating point rounding issues (probably a
non-issue).
Release Notes:
- N/A
It doesn't make sense to have `Pixels: Mul<Pixels, Output = Pixels>` as
the output should be `Pixels^2` (area), so these impls are removed. All
code where these impls were used are improved by instead multiplying by
`f32` or `usize`.
Also adds math op impls that are present for `Pixels` but absent for
`ScaledPixels`. Adds missing `Mul<Pixels> for usize` to both.
Release Notes:
- N/A
This was added in #8343 to make it only visible for tests. #9189 then
made it visible regardless of `test-support`, so the definitions became
identical.
Release Notes:
- N/A
In `layout_excerpt_gutter`, compute max line number length once instead
of for every row
In `max_line_number_width`, use ilog10 instead of converting to floats
and back
Release Notes:
- N/A
Closes#26002
Release Notes:
- Added `multibuffer` key context.
- `cmd-down` and `cmd-shift-down` on Mac now moves to the end of the
last line of a singleton buffer instead of the beginning. In
multibuffers, these now move to the start of the next excerpt.
- Fixed `vim::PreviousSectionEnd` (bound to `[ ]`) to move to the
beginning of the line, matching the behavior of `vim::NextSectionEnd`.
- Added `editor::MoveToStartOfNextExcerpt` and
`editor::MoveToEndOfPreviousExcerpt`.
Access to `cx` will be needed for anything that queries entities. In
this commit this is use of `WorktreeStore::find_search_candidates`. In
the future it will be things like access to LSP / tree-sitter outlines /
etc.
Changes to support access to `cx` from functions provided to the Lua
script:
* Adds a channel of requests that require a `cx`. Work enqueued to this
channel is run on the foreground thread.
* Adds `async` and `send` features to `mlua` crate so that async rust
functions can be used from Lua.
* Changes uses of `Rc<RefCell<...>>` to `Arc<Mutex<...>>` so that the
futures are `Send`.
One benefit of reusing project search logic for search candidates is
that it properly ignores paths.
Release Notes:
- N/A
Covers part of #5129 by adding `MoveToStartOfExcerpt`,
`MoveToEndOfExcerpt`, `SelectToStartOfExcerpt`, and
`SelectToEndOfExcerpt`.
No default linux bindings yet as it's unclear what to use. Currently,
`ctrl-up` / `ctrl-down` scroll up and down by one line (see #13269).
Considering changing the meaning of those.
Mac:
* Previously `cmd-up` and `cmd-down` were `editor::MoveToBeginning` and
`editor::MoveToEnd`. In singleton editors these will behave the same as
before. In multibuffers, they will now step through excerpts instead of
jumping to the beginning / end of the multibuffer.
* `cmd-home` and `cmd-end`, often typed as `cmd-fn-left` and
`cmd-fn-right` are now `editor::MoveToBeginning` and
`editor::MoveToEnd`. This is useful in multibuffers.
Release Notes:
- Mac: `cmd-up` now moves to the previous
multibuffer excerpt start, and `cmd-down` moves to the next multibuffer
excerpt end. Within normal buffers these behave the same as before, moving
to the beginning or end.
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
In Zed the key context almost always has more than 1 entry, so use of
`SmallVec` is just adding overhead.
In Zed while using the editor this typically has more than 8 entries.
Since `ContextEntry` is 48 bytes, if this were made to be a
`SmallVec<[ContextEntry; 10]>` then it would use 480 bytes on the stack,
which to me seems like a lot to be copying. So, instead opting to just
use `Vec`
Release Notes:
- N/A
* Remove unneeded accepted licenses
* Removes use of `workarounds`
- `wasmtime` no longer needed in list
- `ring` now checks the license SHA
* Checks license from `files` instead of from `git`. Execution time ~17s
instead of ~24s
Release Notes:
- N/A
The name `livekit_server` was a bit misleading as it is not a server and
gets built into both the client and server - the server code is in
`collab`.
Release Notes:
- N/A
Rationale for the changes:
* `requires migration` -> `uses some deprecated settings` changed
because really it isn't required by this version of Zed, and I believe
we hope to offer support for deprecated settings and their migration for
a long time.
* Rename of `migration` -> `updated` is because to me, "updated" feels
lighter and more accurate. To me migration has connotations of moving to
a whole new format.
Formatting changes are due to shortening the line causing cargo fmt to
go from not formatting the code to doing so.
Release Notes:
- N/A
---------
Co-authored-by: smit <0xtimsb@gmail.com>
Moves tab whitespace insertion logic out of `AcceptEditPrediction`
handler.
`edit_prediction_requires_modifier` context will now be true when on a
line with leading whitespace, so that `alt-tab` is used to accept
predictions in this case. This way leading indentation can be typed when
edit predictions are visible
Release Notes:
- N/A
Co-authored-by: Ben <ben@zed.dev>
Co-authored-by: Joao <joao@zed.dev>
Was unnecessary to include these in #24596 as they will be available
from the base keymap.
Release Notes:
- N/A
Co-authored-by: Conrad <conrad@zed.dev>
* Overrides the action handler to switch to insert mode after jumps.
* Returns `vim::Tab` to its behavior from before #24418
Release Notes:
- N/A
Co-authored-by: Conrad <conrad@zed.dev>
When working on #24442, I did a project wide replacement of
`AcceptInlineCompletion` with `AcceptEditPrediction`, as I was updating
the branch to mmain and that rename had happened. This also replaced it
in the migrator, causing the migration notification to always pop up on
keymap changes.
Checking if the migration actually changes the text makes it behave
better if this variety of bug happens in the future.
Release Notes:
- N/A
* When an edit prediction is present in non-insertion modes, hide it but
show `tab Jump to edit`.
* Removes discarding of edit predictions when going from insert mode to
normal mode, instead just hide them in non-insertion modes.
* Removes zeta-specific showing of predictions in normal mode. This
behavior was only happening in special cases anyway - where the discard
of completions wasn't happening due to some other thing taking
precedence in `dismiss_menus_and_popups`.
Release Notes:
- N/A
---------
Co-authored-by: Conrad <conrad@zed.dev>
Co-authored-by: Mikayla <mikayla@zed.dev>
`use_key_equivalents` does nothing on linux, as key equivalents are only
supported on mac. While it could be sensible to anticipate support,
right now it is only used in these few spots, so removing it.
Release Notes:
- N/A
Byte size before was 672, now is 56. The `cached` method is only used in
two places, so this was a lot of extra bytes being shuffled around for
every `AnyView` not using this.
Release Notes:
- N/A
Regression in #22644
Unfortunately not a full fix, In the case where a tooltip gets displayed
and then gets occluded after display, it will stick around until the
mouse exits the hover bounds.
Release Notes:
- N/A
Co-authored-by: Ben <ben@zed.dev>
Regressions in #24024:
* `+` was no longer included between modifiers and key
* Multi-character keys like "control" were displayed all lowercase,
whereas before they were all uppercase like "CONTROL". Now they are
capitalized, so "Control".
* Brings back icon for tab key.
Release Notes:
- N/A
Release Notes:
- Fixed some modifier changed events not being present on Linux X11.
This affected things like the project search palette, where holding ctrl
would not cause the split options to appear.
Use of this location info was added in #21758 to help with diagnosing
remote_server panics on drop of tasks on a different thread.
Release Notes:
- N/A
Realized that the logic in #23814 was more than needed, and harder to
maintain. Something like that could make sense if using the tokenizer
and wanting to precisely hit a token limit. However in the case of edit
predictions it's more of a latency+expense vs capability tradeoff, and
so such precision is unnecessary.
Happily this change didn't require much extra work, just copy-modifying
parts of that change was sufficient.
Release Notes:
- N/A
Limits the size of the buffer excerpt and the size of change history.
Release Notes:
- N/A
---------
Co-authored-by: Richard <richard@zed.dev>
Co-authored-by: Joao <joao@zed.dev>
Required using a global `LazyLock<Mutex<AppNotifications>>` instead of a
context global because settings errors first occur before initialization
of the notifications global.
Release Notes:
- Errors in settings file are now reported in UI on startup.
Michael Sloan
