The selector splitting code that's used for JS linting plugins didn't
properly account for selectors being a single character. This can happen
in the case of `*`.
Instead of comparing against the length, we'll now check if the
remaining string portion is not empty, which is more robust. It also
allows us to detect trailing whitespace, which we didn't before.
Instead of hard erroring, we now surface module not found errors as
TypeScript diagnostics (we have yet to show the source code of the
error, but something we can improve over time).
Addresses the review feedback in
https://github.com/denoland/deno/pull/27416 .
- Hoist the buffer max size variable to make it less confusing
- Remove manual AST field counter in favour of an explicit "commit
schema" step which writes the actual field count.
When running selectors for JS linting plugins we would error when
encountering an unknown attribute name:
```js
// selector
Foo[non-existant]
// error
Error: Missing string id: <number>
```
This was caused by using `0` as the invalid marker, but also overloading
`0` with an actual node type. So the fix is to reserve `0` as the
invalid marker and move the property type to the next index.
This PR extracts the core part of
https://github.com/denoland/deno/pull/27203 to make it easier to review
and land in parts.
It contains:
- The JS plugin code the deserializes and walks the buffer
- The Rust portion to serialize SWC to the buffer format (a bunch of
nodes are still todos, but imo these can land anytime later)
- Basic lint plugin types, without the AST node types to make this PR
easier to review
- Added more code comments to explain the format etc.
More fixes and changes will be done in follow-up PRs.
---------
Co-authored-by: Bartek Iwańczuk <biwanczuk@gmail.com>
Implements a QUIC interface, loosely based on the WebTransport API (a
future change could add the WebTransport API, built on top of this one).
[quinn](https://docs.rs/quinn/latest/quinn/) is used for the underlying
QUIC implementation, for a few reasons:
- A cloneable "handle" api which fits quite nicely into deno resources.
- Good collaboration with the rust ecosystem, especially rustls.
- I like it.
<!--
Before submitting a PR, please read https://deno.com/manual/contributing
1. Give the PR a descriptive title.
Examples of good title:
- fix(std/http): Fix race condition in server
- docs(console): Update docstrings
- feat(doc): Handle nested reexports
Examples of bad title:
- fix #7123
- update docs
- fix bugs
2. Ensure there is a related issue and it is referenced in the PR text.
3. Ensure there are tests that cover the changes.
4. Ensure `cargo test` passes.
5. Ensure `./tools/format.js` passes without changing files.
6. Ensure `./tools/lint.js` passes.
7. Open as a draft PR if your work is still in progress. The CI won't
run
all steps, but you can add '[ci]' to a commit message to force it to.
8. If you would like to run the benchmarks on the CI, add the 'ci-bench'
label.
-->
This change fixes the premature exit of worker threads when there are still
remaining pending ops.
This change reuses the idea of #22647 (unref'ing `op_worker_recv_message` in
worker threads if closeOnIdle specified) and uses
`web_worker.has_message_event_listener` check in the opposite way as
#22944. (Now we continue the worker when `has_message_event_listener` is
true instead of stopping it when `has_message_event_listener` is false.
closes #23061
closes #26154
This brings some of the HTTP/2 retry behavior from reqwest to
`ext/fetch`. It will retry very specific HTTP/2 errors once, if the body
is able to be used again.
Closes #27332
This commit fixes `deno task` by checking if the provided
task name actually has a wildcard char ("*").
Previously, if the "--recursive" flag was passed, the task name
was treated as a regex, which lead to a situation where exact task
name resulted in a regex that matched all tasks with the specific
prefix.
This commit fixes it, by checking if the provided task name, is an exact
name, or is it a wildcard match.
Closes https://github.com/denoland/deno/issues/27370
Closes https://github.com/denoland/deno/issues/27401
Closes https://github.com/denoland/deno/issues/27408
Fixes #27038.
Previously, for NPM packages the latest version was the version with the
"latest" tag. For JSR packages, the latest version was the greatest
version that matched a `*` version requirement. Unfortunately, that
doesn't work well with pre-release versions.
This PR changes it so that the latest version is always > the currently
requested version.
For NPM: if "latest" tag > current then "latest" tag; otherwise the
greatest version that is >= current
For JSR: greatest version >= current
This is the most reasonable behavior I could come up with. For example,
```
versions:
2.0.0-beta.2
2.0.0-beta.1
1.0.0 => "latest" tag
with a version req `^2.0.0-beta.1`
previously:
"Update" column => 2.0.0-beta.2
"Latest" column => 1.0.0
now:
"Update" column => 2.0.0-beta.2
"Latest" column => 2.0.0-beta.2
```
Working on loading plugin configuration for
https://github.com/denoland/deno/pull/27203
I encountered a lot of complexity, so did some drive-by cleanups to make
it easier to grok the code and have fewer duplicate names.
Merging as a fix so that LTS gets this as it's a useful diagnostic tool.
The 1MB unique is because we deduplicate files that we store (ex. some
packages have the same file multiple times so we store that once).
I ended up changing the file system implementation to determine
its root directory as the last step of building it instead of being the
first step which makes it much more reliable.
split up otel config into user configurable and runtime configurable
parts. user configurable part is now set via env vars parsed according
to the otel spec. otel is now enabled via `OTEL_DENO=true`, and
`--unstable-otel` only acts as a guard.
Fixes: https://github.com/denoland/deno/issues/27273
