This commit prevents lazy injection of the internal `ErrorHandler` from a destroyed injector, as it would result in another "destroyed injector" error.
PR Close#61886
Similar to `DestroyRef`, this adds the `destroyed` property to
`EnvironmentInjector`. It also has the ability to register callbacks with
`onDestroy`, which throws if `destroyed` is already `true`.
This also omits the bit about whether those callbacks have executed
since I realized the property is set to `true` before executing the
callbacks.
PR Close#61951
Fixes that `getDeferBlocks` wasn't accounting for the case where a component might be injecting `ViewContainerRef`. When that happens, an additional wrapper is introduced that needs to be accounted for when traversing the tree.
Fixes#62047.
PR Close#62156
Adds a field to the directive's metadata tracking whether it has directive dependencies. Knowing this will allow the pipeline to decide whether to produce DOM-only or full instructions.
PR Close#62096
Spec updates are in https://github.com/whatwg/html/pull/10919
For the most part, the updates revolve around the deferred commit
handling (with precommitHandler). Updates to redirect allow more
options. A committed promise now exists on the transition since commits
can be delayed. Tests were made zoneless for easier debugging and
timeouts were reduced.
PR Close#62017
Fixes the following issues with the logic in the unused imports migration that deals with trailing commas:
1. It was generating overlapping text ranges which can break internally.
2. It wasn't handling some cases that produce trailing commas.
PR Close#62118
This commit updates the global error listener to wrap the global ErrorEvent in a new Error with cause
if the error property is undefined.
fixes#62078
PR Close#62081
Adds support for passing in `Binding` objects into `TestBed.createComponent`. This makes it easier to test components by avoiding the need to create a wrapper component. Furthermore, it keeps the behavior consistent between tests and the actual app. For example, given a custom checkbox that looks like this:
```typescript
@Component({
selector: 'my-checkbox',
template: '...',
host: {'[class.checked]': 'isChecked()'}
})
export class MyCheckbox {
isChecked = input(false);
}
```
A test for the `isChecked` input would look something like this:
```typescript
it('should toggle the checked class', () => {
@Component({
imports: [MyCheckbox],
template: '<my-checkbox [isChecked]="isChecked"/>',
})
class Wrapper {
isChecked = false;
}
const fixture = TestBed.createComponent(Wrapper);
const checkbox = fixture.nativeElement.querySelector('my-checkbox');
fixture.detectChanges();
expect(checkbox.classList).not.toContain('checked');
fixture.componentInstance.isChecked = true;
fixture.detectChanges();
expect(checkbox.classList).toContain('checked');
});
```
Whereas with the new API, the test would look like this:
```typescript
it('should toggle the checked class', () => {
const isChecked = signal(false);
const fixture = TestBed.createComponent(MyCheckbox, {
bindings: [inputBinding('isChecked', isChecked)]
});
const checkbox = fixture.nativeElement.querySelector('my-checkbox');
fixture.detectChanges();
expect(checkbox.classList).not.toContain('checked');
isChecked.set(true);
fixture.detectChanges();
expect(checkbox.classList).toContain('checked');
});
```
PR Close#62040
In this commit, we request `APP_ID` outside the `onDestroy` callback because the injector might already be in a destroyed state when the callback runs.
PR Close#61885
Projected nodes were missing ssrId information and were skipping annotating template information, which caused templates to be destroyed and recreated rather than hydrated.
fixes: #50543
PR Close#61989
Due to a bug that is currently in progress of being resolved in the
`rules_js` toolchain (see:
https://github.com/aspect-build/rules_js/issues/362), we were seeing
subtle differences between `main` and PRs/local builds as RBE is a
strict sandbox environment while the normal linux/darwin sandbox isn't
necessarily.
This commit fixes the issue by avoiding the interop targets that don't
bring in the actual transitive node modules.
PR Close#62027
when marked for traversal the reactive context has to be set to null to avoid inheriting the reactive context of the parent component
PR Closes#61662
PR Close#61663
We should remove the `onDestroy` listener once subscription is unsubscribed because components might not be destroyed yet, but they still would capture subscribers.
PR Close#61882
This commit unregisters the `onDestroy` listener when `destroy()` is called on the `ResourceImpl`. This prevents memory leaks and ensures that the resource reference is not captured in the destroy callback after it has already been destroyed.
PR Close#61870
Since `DestroyRef.onDestroy` throws if the `DestroyRef` is already
destroyed, there is a need to be able to tell if it is already destroyed
before attempting to register a callback.
PR Close#61849
Previously this would throw an error on the assertLView when we try to discover the templateLView.
Now this properly returns null for the template consumer and continues discovering other effects on the injector.
PR Close#60772
Adds fix directly for `takeUntilDestroyed` to unsubscribe when already
destroyed instead of putting
synchronous behavior on `DestroyRef.onDestroyed` callback as in #58008fixes#54527
PR Close#61847
Refactor createSignal to return a tuple instead of a signal getter. createSignalTuple will be removed in a follow up pr once createSignalTuple usages in google3 are migrated to createSignal.
PR Close#61705
The observable terminates immediately when `error` is called, and no further emissions or completion notifications occur. Thus, we have to remove the listener in both the `error` and `complete` notifications.
PR Close#61596
Implements a compiler transform that attempts to statically analyze variable names and apply them to usages of signal functions like signal, computed, effect, etc.
PR Close#57348
This commit updates runtime logic to produce an error when there are some `@defer` blocks with `hydrate` triggers, but the incremental hydration is not enabled via `withIncrementalHydration()`. Previously the check was only detecting the case when `withIncrementalHydration()` is present on the server, but missing on the client. With the change in this commit, the check would be performed on the server as well.
PR Close#61741
Follow-up to #61674 where we were leaving behind some whitespace, e.g. `[One, Two, Three]` would turn into `[One ]`. These changes only preserve the whitespace if the node is preceded by a newline.
This wasn't caught by tests, because they were stripping away whitespaces before asserting. I've also reworked the tests to be sensitive to formatting changes.
PR Close#61698
