This PR fixes deviations in our AST format compared to TSEStree. They
are mostly a leftover for when I first started working on it and based
it off of babel instead.
One of the key changes why the changeset is a bit bigger is that
TSEStree uses `undefined` instead of `null` as the empty value for type
nodes. This is likely influenced by `tsc` which use `undefined`
everywhere. The rest of the nodes use `null` though. It's a little
weird, but for now it might be better to align.
(extracted from https://github.com/denoland/deno/pull/27977)
Not sure what our handling of unstable properties in `deno.json` is.
This PR adds it to the config schema.
---------
Signed-off-by: Marvin Hagemeister <marvinhagemeister50@gmail.com>
Co-authored-by: Bartek Iwańczuk <biwanczuk@gmail.com>
This commit adds an unstable lint plugin API.
Plugins are specified in the `deno.json` file under
`lint.plugins` option like so:
```
{
"lint": {
"plugins": [
"./plugins/my-plugin.ts",
"jsr:@deno/lint-plugin1",
"npm:@deno/lint-plugin2"
]
}
}
```
The API is considered unstable and might be subject
to changes in the future.
Plugin API was modelled after ESLint API for the
most part, but there are no guarantees for compatibility.
The AST format exposed to plugins is closely modelled
after the AST that `typescript-eslint` uses.
Lint plugins use the visitor pattern and can add
diagnostics like so:
```
export default {
name: "lint-plugin",
rules: {
"plugin-rule": {
create(context) {
return {
Identifier(node) {
if (node.name === "a") {
context.report({
node,
message: "should be b",
fix(fixer) {
return fixer.replaceText(node, "_b");
},
});
}
},
};
},
},
},
} satisfies Deno.lint.Plugin;
```
Besides reporting errors (diagnostics) plugins can provide
automatic fixes that use text replacement to apply changes.
---------
Co-authored-by: Marvin Hagemeister <marvin@deno.com>
Co-authored-by: David Sherret <dsherret@gmail.com>
interactively select which packages to upgrade. a future improvement
could be to add a way to select the version as well, though not sure how
valuable that would be.
This PR adds the `--permit-no-files` cli options to the `bench`
subcommand. This will cause `deno bench --permit-no-files` to not return
an error when no bench files where found.
This is the release commit being forwarded back to main for 2.1.9
Co-authored-by: bartlomieju <bartlomieju@users.noreply.github.com>
Co-authored-by: Bartek Iwańczuk <biwanczuk@gmail.com>
This is the release commit being forwarded back to main for 2.1.8
Co-authored-by: bartlomieju <bartlomieju@users.noreply.github.com>
Co-authored-by: Bartek Iwańczuk <biwanczuk@gmail.com>
This PR adds support for passing wildcard tasks. All matched tasks are
sorted in case they have dependencies. Tasks already in the dependency
tree will be pruned so that every task only runs once.
```json
{
"tasks": {
"foo-1": "echo 'foo-1'",
"foo-2": "echo 'foo-2'"
}
}
```
```sh
$ deno task "foo-*"
Task foo-1 echo 'foo-1'
foo-1
Task foo-2 echo 'foo-2'
foo-2
```
The changes in the PR look a little bigger than they really are due to
formatting. For the most part, I've only needed to hoist up the task
matching logic.
Closes https://github.com/denoland/deno/issues/26944
Closes https://github.com/denoland/deno/issues/21530
---------
Co-authored-by: Bartek Iwańczuk <biwanczuk@gmail.com>
This makes it so imports of ambient modules (e.g. `$app/environment` in
svelte, any virtual module in vite, or other module provided by a
bundler) don't error in the LSP.
The way this works is that when we request diagnostics from TSC, we also
respond with the list of ambient modules. Then, in the diagnostics code,
we save diagnostics (produced by deno) that may be invalidated as an
ambient module and wait to publish the diagnostics until we've received
the ambient modules from TSC.
The actual ambient modules you get from TSC can contain globs, e.g.
`*.css`. So when we get new ambient modules, we compile them all into a
regex and check erroring imports against that regex. Ambient modules
should change rarely, so in most cases we should be using a pre-compiled
regex, which executes in linear time (wrt the specifier length).
TODO:
- Ideally we should only publish once, right now we publish with the
filtered specifiers and then the TSC ones
- deno check (#27633)
Extracted out of https://github.com/denoland/deno/pull/27838/files
Reduces some allocations by accepting either a pathbuf or url for the
referrer for resolution and returning either a pathbuf or url at the
end, which the caller can then convert into to their preferred state.
This is about 4% faster when still converting the final result to a url
and 6% faster when keeping the result as a path in a benchmark I ran.
This PR updates `deno_lint` which contains a couple of bug fixes for
JSX/React related rules. The react rules now have all a `react-*` prefix
in the name as well.
This commit adds a simple HashMap cache to completion requests.
On a test project the time to compute completions went from 1200ms
to 75ms and the cache contained ~300 entries when 8000 completion
items were produced by the TSC.
This is a short-lived cache and is discarded once the completion
request is finished. In a follow up commits we could make it persist
between requests.
This commit refactors how a snapshot is created for the TypeScript
compiler.
Instead of having 4 ops, only a single op ("op_load") is left. This is
achieved by not creating a "ts.Program" during snapshotting, that during
benchmarking doesn't provide much benefit.
This greatly simplifies build script for the TS snapshot and opens up
way to simplify it even further in follow up PRs.
Allows easily constructing a `DenoResolver` using the exact same logic
that we use in the CLI (useful for dnt and for external bundlers). This
code is then used in the CLI to ensure the logic is always up-to-date.
```rs
use std::rc::Rc;
use deno_resolver:🏭:ResolverFactory;
use deno_resolver:🏭:WorkspaceFactory;
use sys_traits::impls::RealSys;
let sys = RealSys;
let cwd = sys.env_current_dir()?;
let workspace_factory = Rc::new(WorkspaceFactory::new(sys, cwd, Default::default()));
let resolver_factory = ResolverFactory::new(workspace_factory.clone(), Default::default());
let deno_resolver = resolver_factory.deno_resolver().await?;
```
This is a pure refactor, the `99_main_compiler.js` file
was getting out of hand, being over 1500 lines and serving
3 distinct purposes:
- snapshotting
- type-checking
- running LSP
The file was split into:
- 97_ts_host.js
- 98_lsp.js
- 99_main_compiler.js
This commit changes how `deno lint --rules` behaves:
1. All available rules are now printed and rules enabled are marked as
such
2. `deno lint --rules --json` doesn't include markdown documentation
for rules but rather a link to the docs. This should allow us to save
around 400kB of the final `deno` binary size
This is the release commit being forwarded back to main for 2.1.7
Please ensure:
- [x] Everything looks ok in the PR
- [x] The release has been published
To make edits to this PR:
```shell
git fetch upstream forward_v2.1.7 && git checkout -b forward_v2.1.7 upstream/forward_v2.1.7
```
Don't need this PR? Close it.
cc @crowlKats
---------
Co-authored-by: crowlKats <crowlKats@users.noreply.github.com>
Co-authored-by: crowlkats <crowlkats@toaxl.com>
This is achieved by storing CJS export analysis ahead of time in the
executable, which should also improve the performance of `denort` by
this never being done anymore (I'm too lazy atm to bench this, but it
will be significant for some programs).
This slightly degrades the performance of CJS export analysis on
subsequent runs because I changed it to no longer cache in the DENO_DIR
with this PR (denort now properly has no idea about the DENO_DIR). We'll
have to change it to embed this data in the binary and that will also
allow us to get rid of swc in denort (will do that in a follow-up PR).
