This introduces semantic indexing in Zed based on chunking text from
files in the developer's workspace and creating vector embeddings using
an embedding model. As part of this, we've created an embeddings
provider trait that allows us to work with OpenAI, a local Ollama model,
or a Zed hosted embedding.
The semantic index is built by breaking down text for known
(programming) languages into manageable chunks that are smaller than the
max token size. Each chunk is then fed to a language model to create a
high dimensional vector which is then normalized to a unit vector to
allow fast comparison with other vectors with a simple dot product.
Alongside the vector, we store the path of the file and the range within
the document where the vector was sourced from.
Zed will soon grok contextual similarity across different text snippets,
allowing for natural language search beyond keyword matching. This is
being put together both for human-based search as well as providing
results to Large Language Models to allow them to refine how they help
developers.
Remaining todo:
* [x] Change `provider` to `model` within the zed hosted embeddings
database (as its currently a combo of the provider and the model in one
name)
Release Notes:
- N/A
---------
Co-authored-by: Nathan Sobo <nathan@zed.dev>
Co-authored-by: Antonio Scandurra <me@as-cii.com>
Co-authored-by: Conrad Irwin <conrad@zed.dev>
Co-authored-by: Marshall Bowers <elliott.codes@gmail.com>
Co-authored-by: Antonio <antonio@zed.dev>
- Fixed#9729 and #10193
This commit fixes an issue where the string splitting function was
handling characters in the input string improperly. We adjusted the use
of the `take_while` function to calculate the length of the numeric
prefix, rather than directly splitting the string, thus correctly
splitting the string into a numeric prefix part and the remaining part
This PR reverts #6924, for the reasons stated in
https://github.com/zed-industries/zed/pull/6924#issuecomment-2033076300.
It also fixes an issue were the `find_all_references_task_sources`
wasn't cleaned up in all cases.
Release Notes:
- N/A
---------
Co-authored-by: Marshall <marshall@zed.dev>
Co-authored-by: Marshall Bowers <elliott.codes@gmail.com>
This PR extracts the `SemanticVersion` out of `util` and into its own
`SemanticVersion` crate.
This allows for making use of `SemanticVersion` without needing to pull
in some of the heavier dependencies included in the `util` crate.
As part of this the public API for `SemanticVersion` has been tidied up
a bit.
Release Notes:
- N/A
* Store extensions versions' wasm API version in the database
* Share a common struct for extension API responses between collab and
client
* Add wasm API version and schema version to extension API responses
Release Notes:
- N/A
Co-authored-by: Marshall <marshall@zed.dev>
We're doing it. Svelte support is moving into an extension. This PR
fixes some issues that came up along the way.
Notes
* extensions need to be able to retrieve the path the `node` binary
installed by Zed
* previously we were silently swallowing any errors that occurred while
loading a grammar
* npm commands ran by extensions weren't run in the right directory
* Tree-sitter's WASM stdlib didn't support a C function (`strncmp`)
needed by the Svelte parser's external scanner
* the way that LSP installation status was reported was unnecessarily
complex
Release Notes:
- Removed built-in support for the Svelte and Gleam languages, because
full support for those languages is now available via extensions. These
extensions will be suggested for download when you open a `.svelte` or
`.gleam` file.
---------
Co-authored-by: Marshall <marshall@zed.dev>
This upgrades to vscode-eslint 2.4.4 to support flat configs, in
multiple configuration files, ending in `.js`, `.cjs`, `.mjs`.
We changed the code to not use the GitHub release because we actually
don't need the artifacts of the release, we just need the source code,
which we compile anyway.
Fixes#7271.
Release Notes:
- Added support for ESLint flat config files.
([#7271](https://github.com/zed-industries/zed/issues/7271)).
Co-authored-by: Kristján Oddsson <koddsson@gmail.com>
We can convert shell, npm and gulp tasks to a Zed format. Additionally, we convert a subset of task variables that VsCode supports.
Release notes:
- Zed can now load tasks in Visual Studio Code task format
---------
Co-authored-by: Piotr Osiewicz <piotr@zed.dev>
Co-authored-by: Piotr Osiewicz <24362066+osiewicz@users.noreply.github.com>
Co-authored-by: Antonio <antonio@zed.dev>
Resurrected this from some assistant work I did in Spring of 2023.
- [x] Resurrect streaming responses
- [x] Use streaming responses to enable AI via Zed's servers by default
(but preserve API key option for now)
- [x] Simplify protobuf
- [x] Proxy to OpenAI on zed.dev
- [x] Proxy to Gemini on zed.dev
- [x] Improve UX for switching between openAI and google models
- We current disallow cycling when setting a custom model, but we need a
better solution to keep OpenAI models available while testing the google
ones
- [x] Show remaining tokens correctly for Google models
- [x] Remove semantic index
- [x] Delete `ai` crate
- [x] Cloud front so we can ban abuse
- [x] Rate-limiting
- [x] Fix panic when using inline assistant
- [x] Double check the upgraded `AssistantSettings` are
backwards-compatible
- [x] Add hosted LLM interaction behind a `language-models` feature
flag.
Release Notes:
- We are temporarily removing the semantic index in order to redesign it
from scratch.
---------
Co-authored-by: Antonio <antonio@zed.dev>
Co-authored-by: Antonio Scandurra <me@as-cii.com>
Co-authored-by: Thorsten <thorsten@zed.dev>
Co-authored-by: Max <max@zed.dev>
We made a change last week to allow creating files with names. This
means some files have null saved_mtime, which old versions of zed panic
on.
A fix is available in 0.126.3 and above
Release Notes:
- N/A
This fixes#9213 by pinning ESLint to `2.2.20-Insiders` which is the
last known version to work well with Zed.
Once this fix is out, we can take a closer look at upgrading to 2.4.x or
even 3.x once that's out of prerelease.
Release Notes:
- Fixed ESLint integration being broken after Mar 7 2024 due to ESLint
3.0.1 alpha release being pushed.
([#9213](https://github.com/zed-industries/zed/issues/9213)).
To be honest, I am not sure how to use these directories. But since it
is difficult to change these later, if we are going to change them, I
think it is time to do.
Release Notes:
- N/A
This PR moves the Clippy configuration up to the workspace level.
We're using the [`lints`
table](https://doc.rust-lang.org/cargo/reference/workspaces.html#the-lints-table)
to configure the Clippy ruleset in the workspace's `Cargo.toml`.
Each crate in the workspace now has the following in their own
`Cargo.toml` to inherit the lints from the workspace:
```toml
[lints]
workspace = true
```
This allows for configuring rust-analyzer to show Clippy lints in the
editor by using the following configuration in your Zed `settings.json`:
```json
{
"lsp": {
"rust-analyzer": {
"initialization_options": {
"check": {
"command": "clippy"
}
}
}
}
```
Release Notes:
- N/A
This PR fix the "invalid cross-device link" error occurred in linux when
trying to write the settings file atomically, like when click the
"Enable vim mode" checkbox at first start.
```plain
[2024-02-26T22:59:25+08:00 ERROR util] .../zed/crates/settings/src/settings_file.rs:135: Failed to write settings to file "/home/$USER/.config/zed/settings.json"
Caused by:
0: failed to persist temporary file: Invalid cross-device link (os error 18)
1: Invalid cross-device link (os error 18)
```
Currently the `fs::RealFs::atomic_write()` method write to a temp file
created with `NamedTempFile::new()` and then call `persist()` method to
write to the config file path, which actually do a `rename` syscall
under the hood. As the
[issue](https://github.com/Stebalien/tempfile/issues/245) said
> `NamedTempFile::new()` will create a temporary file in your system's
temporary file directory. You need `NamedTempFile::new_in()`.
The temporary file directory in linux is in `/tmp`, which is mounted to
`tmpfs` filesystem, and in most case(all case I guess)
`$HOME/.config/zed` is mounted to a different filesystem. And the
`rename` syscall between different filesystems will return a `EXDEV`
errno, as described in the man page
[rename(2)](https://man7.org/linux/man-pages/man2/renameat2.2.html):
```plain
EXDEV oldpath and newpath are not on the same mounted
filesystem. (Linux permits a filesystem to be mounted at
multiple points, but rename() does not work across
different mount points, even if the same filesystem is
mounted on both.)
```
And as the issue above said, use a different temp dir with
`NamedTempFile::new_in()` for linux platform might be a solution, since
the `rename` syscall provides atomicity.
Release Notes:
- Fix `settings.json` save failed with invalid cross-device link error
in linux
This PR renames the `ZedHttpClient` to `HttpClientWithUrl` to make it
slightly clearer that it still is holding a `dyn HttpClient` as opposed
to being a concrete implementation.
Release Notes:
- N/A
Part of #7108
This PR includes just the static runnables part. We went with **not**
having a dedicated panel for runnables.
This is just a 1st PR out of N, as we want to start exploring the
dynamic runnables front. Still, all that work is going to happen once
this gets merged.
Release Notes:
- Added initial, static Runnables support to Zed. Such runnables are defined in
`runnables.json` file (accessible via `zed: open runnables` action) and
they can be spawned with `runnables: spawn` action.
---------
Co-authored-by: Kirill Bulatov <kirill@zed.dev>
Co-authored-by: Pitor <pitor@zed.dev>
Co-authored-by: Beniamin <beniamin@zagan.be>
Fixes https://github.com/zed-industries/zed/issues/7534 by not requiring
assets for gopls and vscode-eslint langservers — those two are the only
ones in Zed that do not use assets directly when determining langserver
version and retrieving those.
All other servers deal with assets, hence require those to be present.
The problem with https://github.com/tamasfe/taplo/releases is that they
host multiple binary releases in the same release list, so for now the
code works because only the langserver has assets — but as soon as
another release there gets assets, it will break again.
We could filter out those by names also, but they also tend to change
(and can be edited manually), so keeping it as is for now.
Release Notes:
- Fixed gopls language server downloads
([7534](https://github.com/zed-industries/zed/issues/7534))
This PR cleans up the path definitions in `util::paths` following the
Linux merge.
We were using a bunch of target-specific compilation that made these
declarations kind of messy, when really we can limit the conditional
compilation to just the base directories that we use as the basis for
the other directories.
Release Notes:
- N/A
This PR adds the initial support for loading extensions in Zed.
### Extensions Directory
Extensions are loaded from the extensions directory.
The extensions directory has the following structure:
```
extensions/
installed/
extension-a/
grammars/
languages/
extension-b/
themes/
manifest.json
```
The `manifest.json` file is used internally by Zed to keep track of
which extensions are installed. This file should be maintained
automatically, and shouldn't require any direct interaction with it.
Extensions can provide Tree-sitter grammars, languages, and themes.
Release Notes:
- N/A
---------
Co-authored-by: Marshall <marshall@zed.dev>
Reverts zed-industries/zed#7481
This would regress performance because we'd be using the standard
library's hash maps everywhere, so reverting for now.
- Send app version and release stage to collab on connect
- Read the new header on the server
Release Notes:
- Added the ability to collaborate with users on different releases of
Zed.
This PR sorts the dependency lists in our `Cargo.toml` files so that
they are in alphabetical order.
This should make them easier to visually scan when looking for a
dependency.
Apologies in advance for any merge conflicts 🙈
Release Notes:
- N/A
