Previously placeholders were only rendered for dynamic interpolation
expressons in `$localize` tagged strings. But there are also potentially
dynamic values in ICU expressions too, so we need to render these as
placeholders when extracting i18n messages into translation files.
PR Close#38484
This commit updates the code to move generated i18n statements into the `consts` field of
ComponentDef to avoid invoking `$localize` function before component initialization (to better
support runtime translations) and also avoid problems with lazy-loading when i18n defs may not
be present in a chunk where it's referenced.
Prior to this change the i18n statements were generated at the top leve:
```
var I18N_0;
if (typeof ngI18nClosureMode !== "undefined" && ngI18nClosureMode) {
var MSG_X = goog.getMsg(“…”);
I18N_0 = MSG_X;
} else {
I18N_0 = $localize('...');
}
defineComponent({
// ...
template: function App_Template(rf, ctx) {
i0.ɵɵi18n(2, I18N_0);
}
});
```
This commit updates the logic to generate the following code instead:
```
defineComponent({
// ...
consts: function() {
var I18N_0;
if (typeof ngI18nClosureMode !== "undefined" && ngI18nClosureMode) {
var MSG_X = goog.getMsg(“…”);
I18N_0 = MSG_X;
} else {
I18N_0 = $localize('...');
}
return [
I18N_0
];
},
template: function App_Template(rf, ctx) {
i0.ɵɵi18n(2, 0);
}
});
```
Also note that i18n template instructions now refer to the `consts` array using an index
(similar to other template instructions).
PR Close#38404
This commit fixes a regression from "fix(common): ensure
scrollRestoration is writable (#30630)" that caused scrolling to not
happen at all in browsers that do not support scroll restoration. The
issue was that `supportScrollRestoration` was updated to return `false`
if a browser did not have a writable `scrollRestoration`. However, the
previous behavior was that the function would return `true` if
`window.scrollTo` was defined. Every scrolling function in the
`ViewportScroller` used `supportScrollRestoration` and, with the update
in bb88c9fa3d, no scrolling would be
performed if a browser did not have writable `scrollRestoration` but
_did_ have `window.scrollTo`.
Note, that this failure was detected in the saucelabs tests. IE does not
support scroll restoration so IE tests were failing.
PR Close#38468
When removal of one view causes removal of another one from the same
ViewContainerRef it triggers an error with views length calculation. This commit
fixes this bug by removing a view from the list of available views before invoking
actual view removal (which might be recursive and relies on the length of the list
of available views).
Fixes#38201.
PR Close#38317
For a template that contains for example `<span *ngIf="first"></span>`
there's no need to render the `NgIf` guard expression, as the child
scope does not have any type-checking statements, so any narrowing
effect of the guard is not applicable.
This seems like a minor improvement, however it reduces the number of
flow-node antecedents that TypeScript needs to keep into account for
such cases, resulting in an overall reduction of type-checking time.
PR Close#38418
The template type-checker would always generate a directive declaration
even if its type was never used. For example, directives without any
input nor output bindings nor exportAs references don't need the
directive to be declared, as its type would never be used.
This commit makes the `TcbOp`s that are responsible for declaring a
directive as optional, such that they are only executed when requested
from another operation.
PR Close#38418
The template type-checker would generate a statement with a call
expression for all DOM elements in a template of the form:
```
const _t1 = document.createElement("div");
```
Profiling has shown that this is a particularly expensive call to
perform type inference on, as TypeScript needs to perform signature
selection of `Document.createElement` and resolve the exact type from
the `HTMLElementTagNameMap`. However, it can be observed that the
statement by itself does not contribute anything to the type-checking
result if `_t1` is not actually used anywhere, which is only rarely the
case---it requires that the element is referenced by its name from
somewhere else in the template. Consequently, the type-checker can skip
generating this statement altogether for most DOM elements.
The effect of this optimization is significant in several phases:
1. Less type-check code to generate
2. Less type-check code to emit and parse again
3. No expensive type inference to perform for the call expression
The effect on phase 3 is the most significant here, as type-checking is
not currently incremental in the sense that only phases 1 and 2 can
be reused from a prior compilation. The actual type-checking of all
templates in phase 3 needs to be repeated on each incremental
compilation, so any performance gains we achieve here are very
beneficial.
PR Close#38418
This commit replaces the old and slow `ReflectiveInjector` that was
deprecated in v5 with the new `Injector`. Note: This change was only
done in the spec files inside the `aio` folder.
While changing this, it was not possible to directly use `Injector.get`
to get the correct typing for the mocked classes. For example:
```typescript
locationService = injector.get<TestLocationService>(LocationService);
```
Fails with:
> Argument of type 'typeof LocationService' is not assignable to parameter
of type 'Type<TestLocationService> | InjectionToken<TestLocationService> |
AbstractType<TestLocationService>'.
Type 'typeof LocationService' is not assignable to type 'Type<TestLocationService>'.
Property 'searchResult' is missing in type 'LocationService' but required in type
'TestLocationService'.
Therefore, it was necessary to first convert to `unknown` and then to
`TestLocationService`.
```typescript
locationService = injector.get(LocationService) as unknown as TestLocationService;
```
PR Close#38408
We should define ngDevMode to false in Closure, but --define only works in the global scope.
With ngDevMode not being set to false, this size tracking test provides little value but a lot of
headache to continue updating the size.
PR Close#38449
The `@HostListener` functions and lifecycle hooks aren't intended to be public API but
do need to appear in the `.d.ts` files or type checking will break. Adding the
nodoc annotation will correctly hide this function on the docs site.
Again, note that `@internal` cannot be used because the result would be
that the functions then do not appear in the `.d.ts` files. This would
break lifecycle hooks because the class would be seen as not
implementing the interface correctly. This would also break
`HostListener` because the compiled templates would attempt to call the
`onClick` functions, but those would also not appear in the `d.ts` and
would produce errors like "Property 'onClick' does not exist on type 'RouterLinkWithHref'".
PR Close#38448
Fixes an error if a CSS custom property, used inside a host binding, has a
number in its name. The error is thrown because the styling parser only
expects characters from A to Z,dashes, underscores and a handful of other
characters.
Fixes#37292.
PR Close#38432
Previously we added a browser target for `firefox` into the
dev-infra package. It looks like as part of this change, we
accidentally switched the local web testing target to `firefox`.
Web tests are not commonly run locally as we use Domino and
NodeJS tests for primary development. Sometimes though we intend
to run tests in a browser. This would currently work with Firefox
but not on Windows (as Firefox is a noop there in Bazel).
This commit switches the primary browser back to `chromium`. Also
Firefox has been added as a second browser to web testing targets.
This allows us to reduce browsers in the legacy Saucelabs job. i.e.
not running Chrome and Firefox there. This should increase stability
and speed up the legacy job (+ reduced rate limit for Saucelabs).
PR Close#38435
Previously commit message types were provided as part of the ng-dev config in the repository
using the ng-dev toolset. This change removes this configuration expectation and instead
predefines the valid types for commit messages.
Additionally, with this new unified set of types requirements around providing a scope have
been put in place. Scopes are either required, optional or forbidden for a given commit
type.
PR Close#38430
We recently updated the benchpress package to have a more loose
Angular core peer dependency, and less other unused dependencies.
We should make sure to use that in the dev-infra package so that
peer dependencies can be satisified in consumer projects, and so
that less unused dependencies are brought into projects.
PR Close#38440
When a commit message fails validation, rather than throwing out the commit message entirely
the commit message is saved into a draft file and restored on the next commit attempt.
PR Close#38304
Runs the `ng-dev format changed` command whenever the `git commit` command is
run. As all changes which are checked by CI will require this check passing, this
change can prevent needless roundtrips to correct lint/formatting errors. This
automatic formatting can be bypassed with the `--no-verify` flag on the `git commit`
command.
PR Close#38402
Adds an ng-dev formatter option to format all of the staged files. This will can
be used to format only the staged files during the pre-commit hook.
PR Close#38402
Pete Bacon Darwin
Andrew Kushnir
Andrew Scott
Anas Barghoud
waterplea
Ahn
JoostK
Sonu Kapoor
crisbeto
Paul Gschwendtner
Joey Perrott
Sergey Falinsky
Alan Agius
Vlad GURDIGA