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
In this commit, we mark the `SwPush` and `SwUpdate` classes as root providers. As a result, they are no longer statically referenced in the `provideServiceWorker` providers list, which previously forced them to be explicitly bundled into the main file. These classes might never be used—some consumers may use the service worker only for prefetching and caching assets.
Practically speaking, even if a user injects the `SwPush` class without calling `provideServiceWorker()`, it would result in a DI error because the communication channel dependency is not available. There is no practical reason to keep these classes as non-root providers or to reference them explicitly.
Currently, some users work around this by using `patch-package` to modify the service worker code and remove these classes from the providers list.
PR Close#61670
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
Change the command to execute the migration, because the command stated in the previous version results in the error "Cannot find migration 'control-flow-migration' in '@angular/core'.". The new command is more in line with the docs of other migrations.
PR Close#61816
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
Previously, multiple simultaneous calls to `checkForUpdate()` could result in redundant
fetches and hashing of the update manifest, leading to unnecessary network and CPU usage.
This change introduces a mechanism to track an in-progress update check using a cached
promise (`ongoingCheckForUpdate`). Subsequent calls to `checkForUpdate()` while a check
is in progress will return the same promise instead of triggering a new request.
Once the check completes (successfully or not), the cached promise is cleared,
allowing future update checks to proceed normally.
This improves efficiency and prevents overlapping update logic in applications that
may invoke `checkForUpdate()` from multiple sources (e.g. polling, manual triggers).
PR Close#61443
The code editor tooltip exhibits some weird bug when there is a margin applied. The root issue is unknown but at least it improves the current situation.
fixes#60923
PR Close#61782
This update ensures that the service worker can handle range requests, allowing video seeking to work correctly when videos are delivered by the service worker.
PR Close#60029
This ensures that the new optimization rule, and also adev (when we
complete migrating it to `rules_js`) can use the latest compiler-cli
version from HEAD.
PR Close#61810
PR Close#61830
When the expression parser consumes tokens inside a parenthesized expression, it looks for valid tokens until it hits and invalid one or a closing paren. If it finds an invalid token, it reports and error and tries to recover until it finds a closing paren. The problem is that in such cases, it would produce the `ParenthesizedExpression` and continue parsing **from** from the closing paren which would then produce more errors that add noise to the output and result in an incorrect representation of the user's code. E.g. `foo((event.target as HTMLElement).value)` would be recovered to `foo((event.target)).value` instead of `foo((event.target).value)`.
These changes resolve the issue by skipping over the closing paren at the recovery point.
Fixes#61792.
PR Close#61815
Commit 3e70d64 introduced cdk version 20.0.0-rc2 which introduced a change to how the cdk virtual scroll sets some internal state.
Previously in DevTools we were using a computed incorrectly to respond to changes in the directive forest and apply them to the underlying datasource. With the change to the CDK shown above, this incorrect usage caused us to attempt to update underlying signals in the virtual scroll directive while within a computed callback, throwing an error.
This commit corrects our usage by swapping from a computed to an effect, allowing the underlying signals in the scroll directive to be updated without error.
PR Close#61812
When defer blocks have a reference-based trigger without a parameter, we infer it from the placeholder block. This requires some validations like ensuring that there's only one element in the placeholder. The validations are currently implemented at the parser level which can affect tools like linters that need to pass `preserveWhitespaces: true` in order to get accurate source mappings.
These changes move the validations into the template type checker so that they still get flagged, but much later in the process. Moving them over involves a bit more work, because the template type checker also sets `preserveWhitespaces: true`.
Fixes#61725.
PR Close#61747
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
kirjs