Previously the DevTools demo app and browser app had duplicated styles in their respective styles.scss files.
This commit creates a global styles.scss that is imported with sass @use into the demo and browser app styles.scss files. This will prevent any issues where css changes to one are missed in the other. Also reduces duplication of material css theme definitions by consolidating it inone place. The respective styles.scss files for the demo app and browser app continue to exist incase those need environment specific css. For example the browser app requires that height: 100% is set on a document in order to render properly in a browsers devtools tab.
PR Close#49001
RxJS has deprecated the `publishReplay` & `publishLast` operators which will be removed in RxJS 8. `connectable()` should be used instead.
PR Close#48969
We recently switched some of the monitoring e2e tests to Bazel. These
tests should never be cached because they rely on an external URL and
on network access. The URL itself might stay the same for quite a while,
but the underlying site might change based on new deployments. Bazel
only sees the URL and caches the test then. We want to avoid this.
PR Close#49039
Currently as a part of the bootstrapping API migration we comment out the metadata of the root module and instruct users to re-run the module pruning step which can be cumbersome. These changes run the module pruning automatically.
Note that initially I tried to reuse the module pruning logic and to run it against the existing program, but it was problematic, because it was common to have conflicting changes for the same AST nodes.
PR Close#49030
The terser update included various changes, including changes to the
inline optimizations. Constants like for the bloom filter are no longer
inlined multiple times, but instead the constant is preserved and used
with a shortened/mangled name.
Note this also applies to other symbols like `SimpleChanges` too.
This means that such new variables now show up in the symbols goldens.
Concrete examples of inlined & no-longer inlined constants can be
seen in the PR description: https://github.com/angular/angular/pull/49000#issue-1576416106
PR Close#49000
This commit does three things that all related and required to get
rid of `webdriver-manager`:
* Our puppeteer protractor setup in AIO relies on webdriver-manager
because we install a corresponding chromedriver based on the puppeteer
chromium version. We would like to get rid of this brittle setup.
* We don't use `puppeteer` in many places because we manage chromium and
the driver through Bazel. This commit removes the remaining puppeteer
usage and replaces it with the Bazel-managed canonical browser
* We need to migrate the AIO production URL tests to Bazel. These
weren't part of Aspect's migration. This is needed so that we can drop
puppeteer and use the Bazel browser setup.
* Migrates some at-runtime TS `ts-node` test setup to proper idiomatic
Bazel code. Needed because it depends on code that also had to be
migrated to Bazel given the production e2e test Bazel migration (above
points).
Note: The xregexp dependency had to be added to the root project because
`ts_library` does not support compilation deps from `@aio_npm`. This is
something we will fix anyway when we have a more modern toolchain!
PR Close#49025
The Angular devkit updated Terser. Terser landed various changes
related to variable inlining. Most constants are no longer inlined,
but instead the variables are re-used.
This led to a bundle size increase of around ~600B in AIO and most
integration tests.
Here is an example diff:
https://screenshot.googleplex.com/4M8owBYDAqU982T
PR Close#48996
Adds a function that allows for the import resolution for files to be customized in the standalone migration. Externally it's only use is to change `NgForOf` to `NgFor`, but we'll need it internally to deduplicate some Material modules.
Fixes#49006.
PR Close#49022
`setDisabledState` is supposed to be called whenever the disabled state of a control changes, including upon control creation. However, a longstanding bug caused the method to not fire when an *enabled* control was attached. This bug was fixed in v15.
This had a side effect: previously, it was possible to instantiate a reactive form control with `[attr.disabled]=true`, even though the the corresponding control was enabled in the model. (Note that the similar-looking property binding version `[disabled]=true` was always rejected, though.) This resulted in a mismatch between the model and the DOM. Now, because `setDisabledState` is always called, the value in the DOM will be immediately overwritten with the "correct" enabled value.
Users should instead disable the control directly in their model. (There are many ways to do this, such as using the `{value: 'foo', disabled: true}` constructor format, or immediately calling `FooControl.disable()` in `ngOnInit`.)
If this incompatibility is too breaking, you may also opt out using `FormsModule.withConfig` or `ReactiveFormsModule.withConfig` at the time you import it, via the `callSetDisabledState` option.
However, there is an exceptional case: radio buttons. Because Reactive Forms models the entire group of radio buttons as a single `FormControl`, there is no way to control the disabled state for individual radios, so they can no longer be configured as disabled.
In this PR, we have special cased radio buttons to ignore their first call to `setDisabledState` when in `callSetDisabledState: 'always'` mode. This preserves the old behavior.
PR Close#48864
We no longer need webdriver-manager as everything is tested
using Bazel with the Bazel-managed browsers and chrome/gecko drivers.
The drivers and browser binaries are managed as part of dev-infra's
shared browser/driver configuration in `bazel/browsers`.
PR Close#49009
Fixes that the migration was unnecessarily traversing top-level nodes. This was a large performance bottle-neck, because it involves a lot of language service lookups.
PR Close#49010
In some internal migrations the language service was throwing for some file reads which was crashing the migration. These changes add a `try/catch` to avoid interrupting it.
PR Close#49010
For the module pruning and bootstrap API migration steps we depend heavily upon the TypeScript `LanguageService` which ends up being slow on a large project. E.g. in some large internal projects single-file lookups were taking around 30s.
These changes introduce a wrapper around the `LanguageService` that we can use to trick it into not traversing the entire project every time.
PR Close#49010
We're using Node's `path` utilities to remap existing imports and create new ones which can yield paths with forward slash separators.
These changes add some logic to ensure that we only generate forward slashes.
PR Close#48993
Normally having a standalone declaration in the `imports` array is an error and something we handle in the conversion to standalone, but tests can end up in this situation, because apps may have separate tsconfigs for the main app and for tests.
These changes make it so that we move any incorrectly-defined standalone declarations, even if they aren't part of the current migration.
PR Close#48987
For the standalone migration we need to pass a couple of compiler flags which accidentally also overwrote the project's compiler options. These changes extend the options instead.
PR Close#48987
Fixes that we were copying all expressions in the `imports` array of the test to the `imports` of the component, including any potential `ModuleWithProviders`.
Fixes#48971.
PR Close#48987
The first step of the migration updates the tests of the component being migrated, however it also skips any bootstraped declarations. We get back to the these declarations when converting to the standalone bootstrapping APIs, but we weren't updating their tests.
These changes rework some of the logic so that we migrate any remaining tests as a part of the final step.
Relates to #48944.
PR Close#48987
Currently the standalone migration is set up to skip any modules that have a `bootstrap` array with at least one element. This ends up being misleading for small apps who have everything in the root module.
These changes add some logic to only skip the root component.
Fixes#48944.
PR Close#48987
v16 will have a breaking change to the way `UrlTree`s are constructed.
This change is actually a bug fix that makes `UrlTree` creation correct
in more scenarios (see #48508). However, this can affect applications that are
relying on the current incorrect behavior. This commit adds a dev mode
warning when the target of a navigation will change once #48508 is
submitted.
PR Close#48688
In #48898 the `isForwardRef` flag was added to indicate whether a reference should be wrapped in a `forwardRef`. This logic assumed that the node can't be referring to another node within the same file, however from testing it looks like that's not actually the case, because we hit the same code path when an external import to the same symbol exists already.
PR Close#48988
Pawel Kozlowski