This PR improves how we handle completions in buffers with multiple
LSPs.
Context: while working on
https://github.com/zed-industries/zed/issues/19777 with @mgsloan we
noticed that completion triggers coming from language servers are not
tracked properly. Namely, each buffer has `completion_triggers` field
which is read from the configuration of a language server. The problem
is, there can be multiple language servers for a single buffer, in which
case we'd just stick to the one that was registered last.
This PR makes the tracking a bit more fine-grained. We now track not
only what the completion triggers are, but also their origin server id.
Whenever completion triggers are updated, we recreate the completion
triggers set.
Release Notes:
- Fixed completions not triggering when multiple language servers are
used for a single file.
Currently terminal.cursor_shape uses `underline` and `cursor_shape` uses
`underscore`.
This standardizes them so they use the same settings value.
I think `underline` is the more common term and it matches the
terminology used by VSCode, Alacritty, iTerm, etc.
Note the protobuf enum `CursorShape::CursorUnderscore` remains
unchanged.
See also:
- https://github.com/zed-industries/zed/pull/18530
- https://github.com/zed-industries/zed/pull/17572
Release Notes:
- Settings: rename one `cursor_shape` from `underscore` to `underline`
(breaking change).
Per the LSP spec, we should pass .data field of diagnostics into code
action request:
```
/**
* A data entry field that is preserved between a
* `textDocument/publishDiagnostics` notification and
* `textDocument/codeAction` request. *
* @since 3.16.0 */ data?: LSPAny;
```
Release Notes:
- Fixed rare cases where a code action triggered by diagnostic may not
be available for use.
This pull request introduces collaboration for the assistant panel by
turning `Context` into a CRDT. `ContextStore` is responsible for sending
and applying operations, as well as synchronizing missed changes while
the connection was lost.
Contexts are shared on a per-project basis, and only the host can share
them for now. Shared contexts can be accessed via the `History` tab in
the assistant panel.
<img width="1819" alt="image"
src="https://github.com/zed-industries/zed/assets/482957/c7ae46d2-cde3-4b03-b74a-6e9b1555c154">
Please note that this doesn't implement following yet, which is
scheduled for a subsequent pull request.
Release Notes:
- N/A
This reverts commit caed275fbf.
NOTE: this should not be merged until #9668 is on stable and the
`ZedVersion#can_collaborate` is updated to exclude all clients without
that change.
Release Notes:
- N/A
---------
Co-authored-by: Piotr Osiewicz <24362066+osiewicz@users.noreply.github.com>
Once we enable extensions to customize the labels of completions and
symbols, this new structure will allow this to be done with a single
WASM call, instead of one WASM call per completion / symbol.
Release Notes:
- N/A
---------
Co-authored-by: Marshall Bowers <elliott.codes@gmail.com>
Co-authored-by: Marshall <marshall@zed.dev>
This PR adds **internal** ability to run arbitrary language servers via
WebAssembly extensions. The functionality isn't exposed yet - we're just
landing this in this early state because there have been a lot of
changes to the `LspAdapter` trait, and other language server logic.
## Next steps
* Currently, wasm extensions can only define how to *install* and run a
language server, they can't yet implement the other LSP adapter methods,
such as formatting completion labels and workspace symbols.
* We don't have an automatic way to install or develop these types of
extensions
* We don't have a way to package these types of extensions in our
extensions repo, to make them available via our extensions API.
* The Rust extension API crate, `zed-extension-api` has not yet been
published to crates.io, because we still consider the API a work in
progress.
Release Notes:
- N/A
---------
Co-authored-by: Marshall <marshall@zed.dev>
Co-authored-by: Nathan <nathan@zed.dev>
Co-authored-by: Marshall Bowers <elliott.codes@gmail.com>
This PR cleans up a handful of references in doc comments in the
`language` crate so that `rustdoc` will link and display them correctly.
Release Notes:
- N/A
Previously, we would use `Project::serialize_buffer_for_peer` and
`Project::deserialize_buffer` respectively in the host and in the
guest to create a new buffer or just send its ID if the host thought
the buffer had already been sent.
These methods would be called as part of other methods, such as
`Project::open_buffer_by_id` or `Project::open_buffer_for_symbol`.
However, if any of the tasks driving the futures that eventually
called `Project::deserialize_buffer` were dropped after the host
responded with the buffer state but (crucially) before the guest
deserialized it and registered it, there could be a situation where
the host thought the guest had the buffer (thus sending them just the
buffer id) and the guest would wait indefinitely.
Given how crucial this interaction is, this commit switches to creating
remote buffers for peers out of band. The host will push buffers to guests,
who will always refer to buffers via IDs and wait for the host to send them,
as opposed to including the buffer's payload as part of some other operation.
As part of #1405, we changed the way we performed undo and redo to
support combining transactions that were not temporally adjacent for
IME purposes.
We introduced a bug with that release that caused divergence
when performing undo: the bug was caused by only changing the visibility
of fragments whose insertion id was contained in the undo operation. However,
an undo operation also affects deletions which we were mistakenly not
considering. Randomized tests caught this but I guess we didn't run enough
of them.
Now, instead of using these versioned offset ranges, we locate the
fragments associated with a transaction using the transaction's
edit ids. To make this possible, buffers now store a new map called
`insertion_slices`, which lets you look up the ranges of insertions
that were affected by a given edit.
Co-authored-by: Antonio Scandurra <antonio@zed.dev>
* Moving the logic from Rope to text::Buffer makes it easier
to keep the Rope in sync with the fragment tree.
* Removing carriage return characters is lossier, but is much
simpler than incrementally maintaining the invariant that
there are no carriage returns followed by newlines. We may
want to do something smarter in the future.
Co-authored-by: Keith Simmons <keith@zed.dev>
Previously, buffer edits represented empty strings as None
variants of an Option. Now, the edit logic just explicitly
checks for empty strings.
Co-authored-by: Keith Simmons <keith@zed.dev>
