This commit helps to make the section that describes an overview of the syntax of a CLI command stand out from the surrounding text.
Closes#26574
PR Close#43850
The commit which made the `canceledNavigationResolution` property on the `Router`
public did not add the corresponding configuration in the `ExtraOptions`.
https://github.com/angular/angular/commit/3c6b653089837459809a370ebcaf8911c3bab9ed
This was a mistake and is being corrected in this commit. We should not
encourage changing the properties post-setup (i.e.
`inject(Router).canceledNavigationResolution = 'computed'`). This manner
of configuration makes the options non-tree shakeable because we have to
keep both implementations in case the value changes at runtime.
PR Close#43842
Refs #42966.
Previously, a build when emitted one warning and no errors would fail with a non-zero exit code. This is not what users would expect, but had not been an issue before since the compiler did not actually emit any warnings. With upcoming extended template diagnostics and other warnings, this is now a case that needs to be supported. Warnings are printed to `stderr` as before, but `ngc` now exits with code `0` and the build is considered successful.
Implemented this by adding a new `expectedExitCode` parameter to `driveDiagnostics()` which asserts against the real exit code. Most importantly, it does not **require** the the build to pass since any exit code can be given, so it is up to the test to assert this as well as many messages printed to make sure they are acceptable. This is useful for testing warnings and ensuring the build still passes.
PR Close#43673
A rebase of #43507 seems to have inadvertently dropped the note about
the removed `WrappedValue` API from the deprecations page, so this
commit reintroduces the text.
PR Close#43685
improve the aio contributors card by aligning centrally the
contributor's name and adding some padding so that it doesn't
get too close to the card's edges
Note: the name's aligning was already present by wrongly removed in
PR #43355
PR Close#43844
* beforePreactivation hook is unused
* The only place that uses afterPreactivation does not use the arguments
Not to say we won't want to provide hooks similar to this in the future,
but the current state is over-engineered for what it's being used for.
PR Close#43804
It appears that when the tables were changed from HTML to markdown some value types were lost. With this change we add the missing types and remove the commented out HTML content.
PR Close#43839
`Backward compatibility with Lighthouse` and `Backward compatibility with CSS grid` sections are no longer relevant since Internet Explorer is no longer supported.
PR Close#43837
Due to an outdated version of ng-dev running during release the changelog generated did not include a split marker,
it should be added to avoid issues in the future.
PR Close#43830
With this change we update `rollup` and remove the usage of no longer maintained `rollup-plugin-commonjs` and `rollup-plugin-node-resolve` plugins.
PR Close#43737
the tsconfig.json file in aio/content/examples does not seem necessary
anymore as it was only used for tslinting and tslint has been removed
from the examples in PR #43592 and PR #43746
PR Close#43818
The `params` and `queryParams` properties are listed in the deprecations as
properties that "may be deprecated in the future". We should not
document "maybe" deprecations. They are either deprecated or they are
not deprecated. For this case, there is no value in the deprecation
without a plan for how we will remove them. Once that is in place, we
can truly deprecate them.
PR Close#43765
webkit browsers show a cancel X icon in the input search on focus/hover
which allows to clear the content of the input, such icon has small
glitches/imperfections so use a mat-icon instead
(this also adds such icon to non-webkit browsers)
PR Close#43600
Add `<!-- CHANGELOG SPLIT MARKER -->` between changelog entries to match the expected
format of the CHANGELOG.md file for the repository tooling.
PR Close#43799
fix the links present in the animations guide under td elements written as
`<code>[text](link)()</code>` which are not rendered properly
Note:
- the bug was introduced in PR #42885
- the bug is only present for code blocks inside html block elements,
the aforementioned format can be used in the inline html elements without issues
PR Close#43787
Spinning up a Node.js browsers image has little to no overhead when the image is not cached on the host. `install_chrome_libs` however takes takes ~13s everytime.
PR Close#43810
With this change we update Node.js version to 14.17 which is needed because of Eslint 8. Also, we replace the legacy images `circleci/node` with the new images `cimg/node`.
See: https://circleci.com/developer/images/image/cimg/node
PR Close#43810
Re-enables the building of the `src/dev-app` package in the components
repo when the components-repo-unit-test job runs. We previously disabled
the dev-app because it was not compatible with template typechecking of
ngtsc. This should be fixed now and can re-enable it.
This could be valuable for having some actual real test scenarios of
type checking.
PR Close#43798
Currently the TestBed teardown migration is set up in a similar way to all other migrations where we take a `CallExpression`, add a parameter to it, print it, and replace the existing call. The problem is that doing so while preserving the `expression` of the original `CallExpression` can cause comments to be duplicated. This can happen quite frequently, because by default the CLI generates comments before `initTestModule` calls.
To work around it, these changes make the migration more precise by inserting a new parameter or replacing and existing one using string manipulation.
This requires a bit more code, but it's more reliable than the following alternatives:
1. Using `getFullStart` and `getFullWidth` to replace the node. This would work with our current setup, but the problem is that `getFullStart` also includes whitespace and newlines before the leading comment. This can cause us to mess up the user's formatting and figuring out which whitespace to keep and which one to remove is tricky.
2. Recreating the `CallExpression.expression` when constructing the new node. This would also work since it'll drop any existing comments, but the problem is that `CallExpression.expression` can be a wide variety of nodes which we would have to account for. We can't use `getMutableClone`, because it preserves the comments.
Fixes#43739.
PR Close#43776
remove the deprecated tslint from the examples of type systemjs
note: eslint hasn't be applied and linting has been removed entirely
to follow angular's unopinionated view on linting
PR Close#43746
Updates the `karma-sauce-launcher` to the latest version that comes with
the heartbeat implemented. Whenever we create an instance of a browser
in Saucelabs, we need to send a Selenium command every X seconds as otherwise
Saucelabs will terminate the session due to it thinking the session is already
unused. This is especially a problem for Karma unit tests as there is no
selenium command at all, except for the initial one loading the Karma site.
The latest version of the launcher has implemented an interval that
dispatches a noop selenium command every X seconds to notify Saucelabs
that the instance is still active.
The current version of the launcher does not do that. This is likely
the source of some disconnects because after 10-13min, the SL browser
is terminated but Karma still waits some minute after realizing the
browser is actually disconnected.
We (the Angular team), contributed to that version in the past since
the heartbeat was very flaky.. and we removed it. With the new version
a lot has changed, the underlying Selenium client has changed so it's
worth trying again with the heartbeat re-implemented; it worked pretty
well so far in the COMP repo).
PR Close#43659
remove linting from the i18n angular.json example as that is no longer
present in the boilerplace angular.json (and the two should be kept in
sync)
note: the linting was removed from the boilerplate angular.json file in PR #43592
PR Close#43747
In 305d05545a, we accidentally stopped
instantiating the `SwUpdatesService` service, which meant that we were
no longer subscribing to ServiceWorker update events and proactively
checking for updates. We would still gets updates, because the
ServiceWorker automatically checks on each navigation request and when a
new ServiceWorker instance is created, but we would not proactively
fetch updates and also miss `unrecoverable` errors.
This commit fixes this by exposing an `enable()` method from
`SwUpdatesService`, which has to be explicitly called for the service to
be enabled, and ensures that the method is called when `AppComponent` is
initialized.
The commit also gets rid of the `SwUpdatesModule` (whose only purpose
was to provide `SwUpdatesService`) in favor of using `providedIn`.
Fixes#43695
PR Close#43766
Previously, we used to prepend version info to error messages by
updating their `message` property and their `stack` property. However,
due to the fact that on some browsers `error.stack` is lazily computed
based on `error.message` the first time the `stack` property is
accessed, this resulted in prepending the version info twice.
This commit ensures that the version info is only prepended to
`error.stack` once to avoid unnecessary repetition in error messages.
PR Close#43767
Renovate Bot