Part 3/3 of language-service refactoring:
Change all language service APIs to return TS value since Angular LS
will be a proper tsserver plugin. This reduces the need to transform
results among Angular <--> TS <--> LSP.
PR Close#32116
In Angular today, the following pattern works:
```typescript
export class BaseDir {
constructor(@Inject(ViewContainerRef) protected vcr: ViewContainerRef) {}
}
@Directive({
selector: '[child]',
})
export class ChildDir extends BaseDir {
// constructor inherited from BaseDir
}
```
A decorated child class can inherit a constructor from an undecorated base
class, so long as the base class has metadata of its own (for JIT mode).
This pattern works regardless of metadata in AOT.
In Angular Ivy, this pattern does not work: without the @Directive
annotation identifying the base class as a directive, information about its
constructor parameters will not be captured by the Ivy compiler. This is a
result of Ivy's locality principle, which is the basis behind a number of
compilation optimizations.
As a solution, @Directive() without a selector will be interpreted as a
"directive base class" annotation. Such a directive cannot be declared in an
NgModule, but can be inherited from. To implement this, a few changes are
made to the ngc compiler:
* the error for a selector-less directive is now generated when an NgModule
declaring it is processed, not when the directive itself is processed.
* selector-less directives are not tracked along with other directives in
the compiler, preventing other errors (like their absence in an NgModule)
from being generated from them.
PR Close#31379
Adds a new guide that can be used to reproduce failures
reported in the `material-unit-tests` job locally.
The document should live in the framework repository as
the package building scripts are local to the framework
repository.
PR Close#32138
Initially when the `material-unit-tests` job got wired up,
Ivy was not really backwards-compatible and a few bugs caused
test failures when running the Angular Material test suites w/ Ivy.
These bugs got fixed progressively and eventually the test
blocklist became empty. At this point we don't want to regress
in the future and the blocklist should never have new items.
Additionally since we switched the unit-tests job to run against
Angular Material `master` with Bazel, the blocklist is no
longer respected. Therefore we can safely remove the blocklist.
PR Close#32138
The tsserver is not meant to handle HTML files, so there is no point
sending an "open" request. The existing test is wrong because the
quickinfo returns "const name: never", which should be
"(property) WidgetComponent.name"
PR Close#32017
Cleanup the logic in TypeScriptHost as to when langauge service state
should be synchronized with the editor state.
The model employed follows that of tsserver, in which case it is the
caller's responsiblity to synchronize host data before any LS methods
are called.
PR Close#32017
For some reason (on OS/X) this transitive dependency is not being passed
through to the final TS builds that rely on this rule, so the build fails
with a missing file error:
```
The specified path does not exist:
'/.../sandbox/darwin-sandbox/451/execroot/angular/packages/tsconfig-build.json'.
```
PR Close#31943
Updates the `material-unit-tests` job to the latest commit
on the components repository. 097f4335a4e0b6e6b579829ae3a9cffce6292d2b.
This commit ensures that the postinstall script does not run NGC
on schematic code from `@angular/core`. Running NGC on the
generated schematic code can cause unexpected issues as some
migrations import types directly from `@angular/compiler-cli`
while the entry-point is not usable in all cases.
See: https://github.com/angular/angular/issues/29220.
PR Close#31650
Introduces a new migration schematic that follows the given
migration plan: https://hackmd.io/@alx/S1XKqMZeS.
First case: The schematic detects decorated directives which
inherit a constructor. The migration ensures that all base
classes until the class with the explicit constructor are
properly decorated with "@Directive()" or "@Component". In
case one of these classes is not decorated, the schematic
adds the abstract "@Directive()" decorator automatically.
Second case: The schematic detects undecorated declarations
and copies the inherited "@Directive()", "@Component" or
"@Pipe" decorator to the undecorated derived class. This
involves non-trivial import rewriting, identifier aliasing
and AOT metadata serializing
(as decorators are not always part of source files)
PR Close#31650
The $locationShim has onChange listeners to allow for synchronization logic between
AngularJS and Angular. When the AngularJS routing events are emitted first, this can
cause Angular code to be out of sync. Notifying the listeners earlier solves the
problem.
PR Close#32037
PR #29473 changed the docs to use a string as the input value of `formControlName`, as it used to only accept a string.
This has been changed, and `formControlName` now accepts a string or a number, so the example in the docs can use a binding as they used to.
PR Close#30606
This commit relaxes the type of the `formControlName` input to accept both a `string` and a `number`.
Currently, when using a `FormArray`, most templates look like:
```
<div formArrayName="tags">
<div *ngFor="let tag of tagsArray.controls; index as i">
<input [formControlName]="i">
</div>
</div>
```
Here `formControlName` receives a number whereas its input type is a string.
This is fine for VE and `fullTemplateTypeCheck`, but not for Ivy which does a more thorough type checking on inputs with `fullTemplateTypeCheck` enabled and throws `Type 'number' is not assignable to type 'string'`. It is fixable by using `formControlName="{{i}}"` but you have to know the difference between `a="{{b}}"` and `[a]="b"` and change it all over the application codebase. This commit allows the existing code to still type-check.
PR Close#30606
Follow-up to #30993 where we build all Angular packages with
the TypeScript `--strict` flag. The flag improves overall code
health and also helps us catch issues easier.
PR Close#31967
Part 2/3 of language service refactoring:
Now that the language service is a proper tsserver plugin, all LS
interfaces should return TS values. This PR refactors the
ng.getDiagnostics() API to return ts.Diagnostic[] instead of
ng.Diagnostic[].
PR Close#32115
The language service relies on a "context" file that is used as the
canonical "containing file" when performing module resolution.
This file is unnecessary since the language service host's current
directory always default to the location of tsconfig.json for the
project, which would give the correct result.
This refactoring allows us to simplify the "typescript host" and also
removes the need for custom logic to find tsconfig.json.
PR Close#32015
Historically, we've cleaned Ivy commits out of the CHANGELOG because
Ivy was not available except as a preview. Given that Ivy will soon
be the default in 9.0.0, it no longer makes sense to remove the Ivy
commits from the log. This changes the gulp changelog task so that
Ivy commits are included by default.
PR Close#32114
Previously, `validate-commit-message` would treat `fixup! `-prefixed
commits like this:
- It would strip the `fixup! ` prefix.
- It would validate the rest of the commit message header as any other
commit.
However, fixup commits are special in that they need to exactly match an
earlier commit message header (sans the `fixup! ` prefix) in order for
git to treat them correctly. Otherwise, they will not be squashed into
the original commits and will be merged as is. Fixup commits can end up
not matching their original commit for several reasons (e.g. accidental
typo, changing the original commit message, etc.).
This commit prevents invalid fixup commits to pass validation by
ensuring that they match an earlier (unmerged) commit (i.e. a commit
between the current HEAD and the BASE commit).
NOTE: This new behavior is currently not activated in the pre-commit git
hook, that is used to validate commit messages (because the
preceding, unmerged commits are not available there). It _is_
activated in `gulp validate-commit-message`, which is run as part
of the `lint` job on CI and thus will detect invalid commits,
before their getting merged.
PR Close#32023
While `fixup! ` is fine, `squash! ` means that the commit message needs
tweaking, which cannot be done automatically during merging (i.e. it
should be done by the PR author).
Previously, `validate-commit-message` would always allow
`squash! `-prefixed commits, which would cause problems during merging.
This commit changes `validate-commit-message` to make it configurable
whether such commits are allowed and configures the
`gulp validate-commit-message` task, which is run as part of the `lint`
job on CI, to not allow them.
NOTE: This new check is disabled in the pre-commit git hook that is used
to validate commit messages, because these commits might still be
useful during development.
PR Close#32023
There has been a regression where enabling rollup treeshaking causes errors during runtime because it will drop const access which will always evaluate to true or false. However, such `const` in `@angular/core` cannot be dropped because their value is changed when NGCC is run on `@angular/core`
VE
```
const SWITCH_IVY_ENABLED__POST_R3__ = true;
const SWITCH_IVY_ENABLED__PRE_R3__ = false;
const ivyEnabled = SWITCH_IVY_ENABLED__PRE_R3__;
```
Ivy (After NGCC)
```
const SWITCH_IVY_ENABLED__POST_R3__ = true;
const SWITCH_IVY_ENABLED__PRE_R3__ = false;
const ivyEnabled = SWITCH_IVY_ENABLED__POST_R3__;
```
FESM2015
```
load(path) {
/** @type {?} */
const legacyOfflineMode = this._compiler instanceof Compiler;
return legacyOfflineMode ? this.loadFactory(path) : this.loadAndCompile(path);
}
```
ESM2015
```
load(path) {
/** @type {?} */
const legacyOfflineMode = !ivyEnabled && this._compiler instanceof Compiler;
return legacyOfflineMode ? this.loadFactory(path) : this.loadAndCompile(path);
}
```
From the above we can see that `ivyEnabled ` is being treeshaken away when generating the FESM bundle which is causing runtime errors such as `Cannot find module './lazy/lazy.module.ngfactory'` since in Ivy we will always load the factories.
PR Close#32069
Similar to interpolation, we do not want to completely remove whitespace
nodes that are siblings of an expansion.
For example, the following template
```html
<div>
<strong>items left<strong> {count, plural, =1 {item} other {items}}
</div>
```
was being collapsed to
```html
<div><strong>items left<strong>{count, plural, =1 {item} other {items}}</div>
```
which results in the text looking like
```
items left4
```
instead it should be collapsed to
```html
<div><strong>items left<strong> {count, plural, =1 {item} other {items}}</div>
```
which results in the text looking like
```
items left 4
```
---
**Analysis of the code and manual testing has shown that this does not cause
the generated ids to change, so there is no breaking change here.**
PR Close#31962
Previously if only a component template changed then we would know to
rebuild its component source file. But the compilation was incorrect if the
component was part of an NgModule, since we were not capturing the
compilation scope information that had a been acquired from the NgModule
and was not being regenerated since we were not needing to recompile
the NgModule.
Now we register compilation scope information for each component, via the
`ComponentScopeRegistry` interface, so that it is available for incremental
compilation.
The `ComponentDecoratorHandler` now reads the compilation scope from a
`ComponentScopeReader` interface which is implemented as a compound
reader composed of the original `LocalModuleScopeRegistry` and the
`IncrementalState`.
Fixes#31654
PR Close#31932
Keen Yee Liau
Esteban Gehring
Alex Rickabaugh
Paul Gschwendtner
Kapunahele Wong
George Kalpakas
Judy Bogart
Pete Bacon Darwin
Miško Hevery
Pawel Kozlowski
Alison Gale
cexbrayat
Jordan Nelson
Joey Perrott
JiaLiPassion
Kara Erickson
Kristiyan Kostadinov
Alan Agius
Renovate Bot