This PR implements the changes we plan to make to `deno install` in deno
2.0.
- `deno install` without arguments caches dependencies from
`package.json` / `deno.json` and sets up the `node_modules` folder
- `deno install <pkg>` adds the package to the config file (either
`package.json` or `deno.json`), i.e. it aliases `deno add`
- `deno add` can also add deps to `package.json` (this is gated behind
`DENO_FUTURE` due to uncertainty around handling projects with both
`deno.json` and `package.json`)
- `deno install -g <bin>` installs a package as a globally available
binary (the same as `deno install <bin>` in 1.0)
---------
Co-authored-by: Nathan Whitaker <nathan@deno.com>
Fixes the `Debug Failure` errors described in
https://github.com/denoland/deno/issues/23643#issuecomment-2094552765 .
The issue here was that we were passing diagnostic codes as strings but
TSC expects the codes to be numbers. This resulted in some quick fixes
not working (as illustrated by the test added here which fails before
this PR).
The first commit is the actual fix. The rest are just test related.
A bunch of small things, mostly around timing and making sure the
jupyter kernel is actually running and ready to respond to requests. I
reproduced the flakiness by running a script to run a bunch of instances
of the test in parallel, where I could get failures consistently. After
this PR, I can't reproduce the flakiness locally which hopefully means
that applies to CI as well
Allows writing named sub-tests. These are:
1. Filterable on the command line via `cargo test ...`
2. Run in parallel
3. Use a fresh temp and deno dir for each test (unlike steps)
This commit updates our testing npm registry to handle
additional `@denotest2` scope in addition to `@denotest`
scope. I might have to update it further in the future to handle
additional scopes, but it's good enough for now.
Part of #22607 (probably closes it, but I haven't done thorough testing)
Makes it so that `require.resolve` with `paths` specified will fallback
to using the global cache when the paths can't be found when using a
global cache (not when using a node_modules folder)
This correctly creates the `AbortSignal` regardless of when we request
it. If the signal is requested after the request has completed, the
signal is created in the aborted state.
Using GC counts, we can see a reduction in object creation:
This PR: 440
deno 1.42.4: 1650
deno 1.43.0+b02ffec: 874
This PR adds private `[REF]()` and `[UNREF]()` methods to Stdin class,
and call them from Node.js polyfill layer (`TTY` class). This enables
`process.stdin.unref()` and `process.stdin.ref()` for the case when
stdin is terminal.
closes #21796
By default, `deno serve` will assign port 8000 (like `Deno.serve`).
Users may choose a different port using `--port`.
`deno serve /tmp/file.ts`
`server.ts`:
```ts
export default {
fetch(req) {
return new Response("hello world!\n");
},
};
```
When the response has been successfully send, we abort the
`Request.signal` property to indicate that all resources associated with
this transaction may be torn down.
Most common argument to `env` option for `worker_threads.Worker` will be
`process.env`.
In Deno `process.env` is a `Proxy` which can't be cloned using
structured clone algorithm.
So to be safe, I'm creating a copy of actual object before it's sent to
the worker thread.
Ref #23522
This commit adds a "private npm registry" to the test server. This
registry requires to send an appropriate Authorization header.
Towards https://github.com/denoland/deno/issues/16105
This commit changes the workspace support to provide all workspace
members to be available as imports based on their names and versions.
Closes https://github.com/denoland/deno/issues/23343
<!--
Before submitting a PR, please read
https://docs.deno.com/runtime/manual/references/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 PR wires up a new `jsxPrecompileSkipElements` option in
`compilerOptions` that can be used to exempt a list of elements from
being precompiled with the `precompile` JSX transform.
The actual handling of `$projectChanged` is quick, but JS requests are
not. The cleared caches only get repopulated on the next actual request,
so just batch the change notification in with the next actual request.
No significant difference in benchmarks on my machine, but this speeds
up `did_change` handling and reduces our total number of JS requests (in
addition to coalescing multiple JS change notifs into one).
Embedders may have special requirements around file opening, so we add a
new `check_open` permission check that is called as part of the file
open process.
Adds an `addr` field to `HttpServer` to simplify the pattern
`Deno.serve({ onListen({ port } => listenPort = port })`. This becomes:
`const server = Deno.serve({}); port = server.addr.port`.
Changes:
- Refactors `serve` overloads to split TLS out (in preparation for
landing a place for the TLS SNI information)
- Adds an `addr` field to `HttpServer` that matches the `addr` field of
the corresponding `Deno.Listener`s.
It's not clear to me how these tests worked correctly on CI,
but they were failing hard locally because of two problems:
- missing env var that tests URL for fake npm registry
- trying to run a directory that contains native Node.js tests that
require a special harness
Landing work from #21903, plus fixing a node compat bug.
We were always sending the HTTP/2 ALPN on TLS connections which might
confuse upstream servers.
Changes:
- Configure HTTP/2 ALPN when making the TLS connection from the HTTP/2
code
- Read the `ALPNProtocols` property from the TLS connection options
rather than the deno `alpnProtocols` field
- Add tests
Prereq for landing Deno.serveHttp on Deno.serve: removing older HTTP
servers from the codebase.
This PR enables V8 code cache for ES modules and for `require` scripts
through `op_eval_context`. Code cache artifacts are transparently stored
and fetched using sqlite db and are passed to V8. `--no-code-cache` can
be used to disable.
---------
Co-authored-by: Bartek Iwańczuk <biwanczuk@gmail.com>
It's best that this only gets merged with the latest version of the
suite, so there's little difference between the `ci` and `wpt_epoch`
workflows. This should make troubleshooting easier.
This allows people to use imports like:
```ts
import "./app.css";
```
...with `deno check` in systems where there's a bundle step (ex. Vite).
This will still error when using it with `deno run` or if the referenced
file does not exist.
See test cases for behaviour.
This PR adds a benchmark intended to measure how the LSP handles larger
repos, as well as its performance on a more realistic workload.
The repo being benchmarked is
[deco-cx/apps](https://github.com/deco-cx/apps) which has been vendored
along with its dependencies. It's included as a git submodule as its
fairly large. The LSP requests used in the benchmark are the actual
requests sent by VSCode as I opened, modified, and navigated around a
file (to simulate an actual user interaction).
The main motivation is to have a more realistic benchmark that measures
how we do with a large number of files and dependencies. The
improvements made from 1.42 to 1.42.3 mostly improved performance with
larger repos, so none of our existing benchmarks showed an improvement.
Here are the results for the changes made from 1.42 to 1.42.3 (the new
benchmark is the last one listed):
**1.42.0**
```test
Starting Deno benchmark
-> Start benchmarking lsp
- Simple Startup/Shutdown
(10 runs, mean: 379ms)
- Big Document/Several Edits
(5 runs, mean: 1142ms)
- Find/Replace
(10 runs, mean: 51ms)
- Code Lens
(10 runs, mean: 443ms)
- deco-cx/apps Multiple Edits + Navigation
(5 runs, mean: 25121ms)
<- End benchmarking lsp
```
**1.42.3**
```text
Starting Deno benchmark
-> Start benchmarking lsp
- Simple Startup/Shutdown
(10 runs, mean: 383ms)
- Big Document/Several Edits
(5 runs, mean: 1135ms)
- Find/Replace
(10 runs, mean: 55ms)
- Code Lens
(10 runs, mean: 440ms)
- deco-cx/apps Multiple Edits + Navigation
(5 runs, mean: 11675ms)
<- End benchmarking lsp
```
Closes https://github.com/denoland/deno/issues/23362
Previously we were panicking if there was a pending read on a
port and `receiveMessageOnPort` was called. This is now fixed
by cancelling the pending read, trying to read a message and
resuming reading in a loop.
…faces (#23296)"
This reverts commit e190acbfa8.
Reverting because it broke stable API type declarations. We will reland
it for v1.43 with updated interfaces
Currently we evict a lot of the caches on the JS side of things on every
request, namely script versions, script file names, and compiler
settings (as of #23283, it's not quite every request but it's still
unnecessarily often).
This PR reports changes to the JS side, so that it can evict exactly the
caches that it needs too. We might want to do some batching in the
future so as not to do 1 request per change.
Removes the certificate options from all the interfaces and replaces
them with a new `TlsCertifiedKeyOptions`. This allows us to centralize
the documentation for TLS key management for both client and server, and
will allow us to add key object support in the future.
Also adds an option `keyFormat` field to the cert/key that must be
omitted or set to `pem`. This will allow us to load other format keys in
the future `der`, `pfx`, etc.
In a future PR, we will add a way to load a certified key object, and we
will add another option to `TlsCertifiedKeyOptions` like so:
```ts
export interface TlsCertifiedKeyOptions =
| TlsCertifiedKeyPem
| TlsCertifiedKeyFromFile
| TlsCertifiedKeyConnectTls
| { key: Deno.CertifiedKey }
```
Changes:
- Implements a TCP socket listener that will allow for round-robin
load-balancing in-process.
- Cleans up the raw networking code to make it easier to work with.
This PR follows this fix (https://github.com/nodejs/node/pull/52005) in
Node.js.
Stream's construct callback happens one tick earlier by this change, and
it prevents the reordering of the first few chunks in
`node:stream.Writable`
closes #20284
I'm unsure whether we're planning to make the `Deno.FsFile` constructor
illegal or remove `FsFile` from the `Deno.*` namspace in Deno 2. Either
way, this PR works towards the former. I'll create a superceding PR if
the latter is planned instead.
Towards #23089
This functionality was broken. The series of events was:
1. Load the npm resolution from the lockfile.
2. Discover only a subset of the specifiers in the documents.
3. Clear the npm snapshot.
4. Redo npm resolution with the new specifiers (~500ms).
What this now does:
1. Load the npm resolution from the lockfile.
2. Discover only a subset of the specifiers in the documents and take
into account the specifiers from the lockfile.
3. Do not redo resolution (~1ms).
MessagePort if directly assigned to workerData property instead of
embedding it in an object then it is not patched to a NodeMessagePort.
This commit fixes the bug.
To avoid the risk of port collisions during tests, we listen on port 0
and use that for both ends of the connections (for any tests we run in
this file).
Fixes #23179.
Fixes #22454.
Enables passing `{tokens: true}` to `parseArgs` and setting default
values for options.
With this PR, the observable framework works with deno out of the box
(no unstable flags needed).
The existing code was basically copied straight from node, so this PR
mostly just updates that (out of date) vendored code. Also fixes some
issues with error exports (before this PR, in certain error cases we
were attempting to construct error classes that weren't actually in
scope).
The last change (in the second commit) adds a small hack so that we
actually exercise the `test-parse-args.js` node_compat test, previously
it was reported as passing though it should have failed. That test now
passes.
---------
Co-authored-by: Bartek Iwańczuk <biwanczuk@gmail.com>
When `DENO_FUTURE=1` env var is present, then BYONM
("bring your own node_modules") is enabled by default.
That means that is there's a `package.json` present, users
are expected to explicitly install dependencies from that file.
Towards https://github.com/denoland/deno/issues/23151
The `tools/node_compat/node` submodule has been moved to
`tests/node_compat/runner/suite` and the remaining files within
`tools/node_compat` to `tests/node_compat/runner`.
Most of the changes are of the header within `tests/node_compat/test`
files. The `setup` and `test` tasks within `tests/node_comapt` execute
successfully.
Towards #22525
CC @mmastrac
The permission prompt doesn't wait for quiescent input, so someone
pasting a large text file into the console may end up losing the prompt.
We enforce a minimum human delay and wait for a 100ms quiescent period
before we write and accept prompt input to avoid this problem.
This does require adding a human delay in all prompt tests, but that's
pretty straightforward. I rewrote the locked stdout/stderr test while I
was in here.
Unused locals and parameters don't make sense to surface in remote
modules. Additionally, fast check can cause these kind of diagnostics
when publishing, so they should be ignored.
Closes #22959
There's a TOCTOU issue that can happen when selecting unused ports for
the server to use (we get assigned an unused port by the OS, and between
then and when the server actually binds to the port another test steals
it). Improve this by checking if the server existed soon after setup,
and if so we retry starting it. Client connection can also fail
spuriously (in local testing) so added a retry mechanism.
This also fixes a hang, where if the server exited (almost always due to
the issue described above) before we connected to it, attempting to
connect our client ZMQ sockets to it would just hang. To resolve this, I
added a timeout so we can't wait forever.
