This commit upgrades hyper-util to 0.1.10, the current latest version.
This also removes exact version specifier in hyper-util to allow library
consumers (e.g. one who wants to use `deno_fetch` as a dependency) to
decide which version to use as long as its version is 0.1.z where z >=
10.
Specifically, hyper-util 0.1.10 is required by Deno Deploy to tweak
`http2_max_header_list_size` (see [hyper-util v0.1.10
changelog](https://github.com/hyperium/hyper-util/releases/tag/v0.1.10))
This reverts commit 3ddbea62c2.
Looks like VMs are now slightly different than before and require
additional setup. Reverting for now to unblock `main` branch.
This commit changes "node:http" module to add support
for the "createConnection" option when the "request()"
API is called.
Closes https://github.com/denoland/deno/issues/19507
---------
Signed-off-by: Yoshiya Hinosawa <stibium121@gmail.com>
Signed-off-by: Satya Rohith <me@satyarohith.com>
Co-authored-by: Yoshiya Hinosawa <stibium121@gmail.com>
Co-authored-by: Bartek Iwańczuk <biwanczuk@gmail.com>
Co-authored-by: crowlkats <crowlkats@toaxl.com>
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
Currently deno eagerly caches all npm packages in the workspace's npm
resolution. So, for instance, running a file `foo.ts` that imports
`npm:chalk` will also install all dependencies listed in `package.json`
and all `npm` dependencies listed in the lockfile.
This PR refactors things to give more control over when and what npm
packages are automatically cached while building the module graph.
After this PR, by default the current behavior is unchanged _except_ for
`deno install --entrypoint`, which will only cache npm packages used by
the given entrypoint. For the other subcommands, this behavior can be
enabled with `--unstable-npm-lazy-caching`
Fixes #25782.
---------
Signed-off-by: Nathan Whitaker <17734409+nathanwhit@users.noreply.github.com>
Co-authored-by: Luca Casonato <hello@lcas.dev>
This commit provides schema files for lint rules and lint tags
in this repo instead of pulling them from `deno_lint` repository.
A unit test was added to ensure all available rules are listed
in the schema file. A unit test for tags can be done once
https://github.com/denoland/deno/pull/27162 lands.
This commit changes "deno outdated" subcommand to
error out if run in a directory that has no config file
(including parent directories). This matches
"pnpm" behavior.
Also added tests for filtering that yields no results,
to ensure that it exists cleanly, that also matches "pnpm"
behavior.
Closes https://github.com/denoland/deno/issues/27287
---------
Signed-off-by: Bartek Iwańczuk <biwanczuk@gmail.com>
Co-authored-by: David Sherret <dsherret@users.noreply.github.com>
Fixes https://github.com/denoland/deno/issues/27289
We exported these but forgot to add them to the list of builtins used by
the resolver, so we weren't resolving bare imports of some modules (e.g.
`"_http_common"`)
Also adds a missing export of `HTTPParser` from `_http_common`
When using the `eval` option on Node's `worker_threads` the code is
passed as a `data:` URL. But we didn't encode the actual code for that,
which lead to syntax errors when including characters not allowed in an
URL.
Fixes a part of https://github.com/denoland/deno/issues/27167
This was doing an allocation for reparsing the specifier. Might as well
do `.join` here and it means I can extract out this file fetcher code to
deno_cache_dir more easily.
`SafeMap` treats its argument as an object with a "length" and index
properties, rather than a generic iterator, so every time we cloned it,
it was dropping all the data.
This PR enables node compat test cases found passing by using the tool
added in #27122
The percentage of passing test case increases from 16.16% to 30.43% by
this change.
