This is the final patch to migrate the Angular styling code to have a
smaller instruction set in preparation for the runtime refactor. All
styling-related instructions now work both in template and hostBindings
functions and do not use `element` as a prefix for their names:
BEFORE:
elementStyling()
elementStyleProp()
elementClassProp()
elementStyleMap()
elementClassMap()
elementStylingApply()
AFTER:
styling()
styleProp()
classProp()
styleMap()
classMap()
stylingApply()
PR Close#30318
This patch is one commit of many patches that will unify all styling instructions
across both template-level bindings and host-level bindings. This patch in particular
removes the `elementIndex` param because it is already set prior to each styling
instruction via the `select(n)` instruction.
PR Close#30313
The `Δ` caused issue with other infrastructure, and we are temporarily
changing it to `ɵɵ`.
This commit also patches ts_api_guardian_test and AIO to understand `ɵɵ`.
PR Close#29850
* Updates the instructions on how to run the benchmark tests.
* Removes the unused `favicon.ico` file and the corresponding Bazel filegroup
PR Close#28645
The web_worker images example is currently not really usable
because the rendered button that can be used to upload
an "image" to the demo is currently not working. This is because
the HTML markup for the `file-field` is not matching what `materialize-css`
expects. See: https://materializecss.com/text-inputs.html
PR Close#28562
Currently all playground examples are built with NGC, and most
of the HTML resources are automatically inlined. Surprisingly NGC
is able to resolve the relative component assets even though these
aren't specified in the `assets`. This seems to work because NGC
resolves the files in the execroot where the files are present
(if Bazel doesn't use sandboxing).
Issue is tracked with TOOL-667
PR Close#28562
The `web_workers/images` example is not being tested by any e2e
spec and therefore it's technically not necessary to fix that it uses
external resources, though in order to ensure that the Bazel builds
are hermetic and that we can eventually add e2e specs for the
web_worker/image example, we should avoid any use of external
resources.
We remove the `web-animations` polyfill in the `web_workers/animations`
example because we should try to vendor as few as possible deps. Also
the animations API is already supported by browsers we run the e2e tests
against (note here: `web_workers/animations` is currently also disabled)
PR Close#28562
The "tree/polymer_leaves" benchmark has no benchmark tests, nor do we install
Polymer anywhere. Polymer was previously installed through `bower`, but since
we removed bower, there is no easy way to run this benchmark with Polymer.
Considering that there are no benchmark tests, nor we have an easy way to
install/vendor Polymer, we should just remove this benchmark app.
This is also based on the assumption that we want to remove the
Polymer benchmarks anyway: see: 8a05199fb9
PR Close#28568
We need to support rendering in contexts like web workers where
nodes are emulated and properties may not be set directly. This
commit gates property validation checks to environments that have
real Node objects.
FW-1043 #resolve
PR Close#28610
Note that this allows Angular to depend on the entirety of the ES2015 API, not just our restricted subset.
This change is needed because our copy of the subset was out-of-date, and prevents us using ES2015 target in dev mode.
This is a subset of #27738
PR Close#28570
Note that this allows Angular to depend on the entirety of the ES2015 API, not just our restricted subset.
This change is needed because our copy of the subset was out-of-date, and prevents us using ES2015 target in dev mode.
This is a subset of https://github.com/angular/angular/pull/27738
PR Close#28134
Currently the "routing" playground example fails the e2e tests
because it tries to load the OpenSans font using an external
HTTP request. External http requests are not allowed (unless
explicitly enabled) within Bazel in order to ensure that
all targets are built and tested in a hermetic way.
In order to work around this issue in a Bazel idiomatic way,
we just vendor the fonts in the "third_party" folder. Note
that we can technically also enable internet for the RBE
host platform, but it's not a best practice for hermeticity.
The following syntax would allow us to enable internet for
RBE (stated here for tracking)
```
properties: {
name: "dockerNetwork"
value: "standard"
}
```
PR Close#28490
With ed1ba88ffd9d0fc266808413fa517e7a31943bc8 we switched the
examples to run with Bazel. This means that we can now also run the
e2e tests for these examples against Ivy. All playground e2e tests,
**except** the `web_worker` examples, successfully run with Ivy.
The failing webworker e2e tests have been temporarily disabled with
`fixmeIvy` and need to be investigated in a follow-up.
PR Close#28490
The `upgrade` playground example has been disabled for a
long time because Protractor initially didn't support running
hybrid apps. Now that we use a more recent version of Protractor
that handles hybrid apps (also automatically), we can re-enable
this long-standing disabled test.
Additionally the e2e test logic was outdated and failed because a
CSS selector did not match the template of the upgrade example.
With this change, the CSS selector has been updated to match the
example's template, and also the test has been updated slightly
to also ensure that content projection works.
PR Close#28490
Currently when someone serves the "person_management" playground
example, there will be runtime exceptions by `@angular/forms` if
someone clicks on one of the two buttons rendered in the example.
This happens because the example is outdated and the input elements
using `ngModel` do not specify a proper "name" while being inside of
a `<form>` element. A name is required inside of a form. The failure
is not specific to Ivy and is not covered by any test because the e2e
tests for this example are just asserting that the page properly loads
(the error only shows up one of the buttons has been clicked)
This is the reason why these errors were never visibile to the e2e tests.
Though in order to make this example work, we should this fix these failures
so that the example can work as expected.
```
FullNameComponent.html:7 ERROR Error: If ngModel is used within a form tag, either the name attribute must be set or the form
control must be defined as 'standalone' in ngModelOptions.
Example 1: <input [(ngModel)]="person.firstName" name="first">
Example 2: <input [(ngModel)]="person.firstName" [ngModelOptions]="{standalone: true}">
at Function.TemplateDrivenErrors.missingNameException (template_driven_errors.ts:40)
at NgModel._checkName (ng_model.ts:319)
at NgModel._checkForErrors (ng_model.ts:302)
at NgModel.ngOnChanges (ng_model.ts:215)
at Object.checkAndUpdateDirectiveInline (provider.ts:208)
at checkAndUpdateNodeInline (view.ts:429)
at checkAndUpdateNode (view.ts:389)
at debugCheckAndUpdateNode (services.ts:431)
at debugCheckDirectivesFn (services.ts:392)
at Object.eval [as updateDirectives] (FullNameComponent.html:7)
```
PR Close#28490
Currently all playground examples are built with `tsc`
and served with the `gulp serve` task. In order to be able
to test these examples easily with Ivy, we now build and
serve the examples using Bazel. This allows us to expand our
Ivy test coverage and additionally it allows us to move forward
with the overall Bazel migration. Building & serving individual
examples is now very easy and doesn't involve building everything
inside of `/modules`.
PR Close#28490
Currently we depend on the "rules_webtesting" version that is
installed by "rules_typescript//:package.bzl". This version of
the webtesting rules comes with a very old version of Chromium
and the `chromedriver` that does not support capturing console
errors properly (with stack traces). Since we have a few e2e
tests that depend on console output (e.g. playground/src/source-map),
we need to make sure that these tests can pass upon Bazel
migration.
PR Close#28490
the polymer benchmarks are super old and not relevant any more
and these benchmarks were the only reason why we needed bower at all
so long, bower. thanks for all the fish.
PR Close#27931
This target fails with odd error that we need to investigate:
[01:19:32] I/direct - Using ChromeDriver directly...
[01:19:32] E/direct - Error code: 135
[01:19:32] E/direct - Error message: Could not find chromedriver at /home/circleci/.cache/bazel/_bazel_circleci/9ce5c2144ecf75d11717c0aa41e45a8d/external/ngdeps/node_modules/webdriver-manager/selenium/chromedriver_2.41. Run 'webdriver-manager update' to download binaries.
[01:19:32] E/direct - Error: Could not find chromedriver at /home/circleci/.cache/bazel/_bazel_circleci/9ce5c2144ecf75d11717c0aa41e45a8d/external/ngdeps/node_modules/webdriver-manager/selenium/chromedriver_2.41. Run 'webdriver-manager update' to download binaries.
PR Close#26735
These tests were previously not running on CI so they have always been broken,
or got broken just recently :-(.
test(ivy): mark failing test targets with fixme-ivy-jit and fixme-ivy-local tags
PR Close#26735
This commit also removes the extra jasminewd2 typings, since the changes
have been merged in the official typings with
DefinitelyTyped/DefinitelyTyped#28957.
PR Close#26139
Previously in Ivy, metadata for directives/components/modules/etc was
carried in .d.ts files inside type information encoded on the
DirectiveDef, ComponentDef, NgModuleDef, etc types of Ivy definition
fields. This works well, but has the side effect of complicating Ivy's
runtime code as these extra generic type parameters had to be specified
as <any> throughout the codebase. *DefInternal types were introduced
previously to mitigate this issue, but that's the wrong way to solve
the problem.
This commit returns *Def types to their original form, with no metadata
attached. Instead, new *DefWithMeta types are introduced that alias the
plain definition types and add extra generic parameters. This way the
only code that needs to deal with the extra metadata parameters is the
compiler code that reads and writes them - the existence of this metadata
is transparent to the runtime, as it should be.
PR Close#26203
With these changes, the types are a little stricter now and also not
compatible with Protractor's jasmine-like syntax. So, we have to also
use `@types/jasminewd2` for e2e tests (but not for non-e2e tests).
I also had to "augment" `@types/jasminewd2`, because the latest
typings from [DefinitelyTyped][1] do not reflect the fact that the
`jasminewd2` version (v2.1.0) currently used by Protractor supports
passing a `done` callback to a spec.
[1]: 566e039485/types/jasminewd2/index.d.ts (L9-L15)Fixes#23952Closes#24733
PR Close#19904
This change supports compilation of components, directives, and modules
within ngtsc. Support is not complete, but is enough to compile and test
//packages/core/test/bundling/todo in full AOT mode. Code size benefits
are not yet achieved as //packages/core itself does not get compiled, and
some decorators (e.g. @Input) are not stripped, leading to unwanted code
being retained by the tree-shaker. This will be improved in future commits.
PR Close#24427
Ivy definition looks something like this:
```
class MyService {
static ngInjectableDef = defineInjectable({
…
});
}
```
Here the argument to `defineInjectable` is well known public contract which needs
to be honored in backward compatible way between versions. The type of the
return value of `defineInjectable` on the other hand is private and can change
shape drastically between versions without effecting backwards compatibility of
libraries publish to NPM. To our users it is effectively an `OpaqueToken`.
By prefixing the type with `ɵ` we are communicating the the outside world that
the value is not public API and is subject to change without backward compatibility.
PR Close#23371
Remove `containerRefreshStart` and `containerRefreshEnd` instruction
from the output.
Generate directives as a list in `componentDef` rather than inline into
instructions. This is consistent in making selector resolution runtime
so that translation of templates can follow locality.
PR Close#22921
BREAKING CHANGE:
The `<template>` tag was deprecated in Angular v4 to avoid collisions (i.e. when
using Web Components).
This commit removes support for `<template>`. `<ng-template>` should be used
instead.
BEFORE:
<!-- html template -->
<template>some template content</template>
# tsconfig.json
{
# ...
"angularCompilerOptions": {
# ...
# This option is no more supported and will have no effect
"enableLegacyTemplate": [true|false]
}
}
AFTER:
<!-- html template -->
<ng-template>some template content</ng-template>
PR Close#22783
Adds a stub for `elementStyle` and `elementClass` instruction
with a canonical spec for the compiler. The spec shows the the
compiler should be using `elementStyle` and `elementClass` instruction
in place of `[class]` and `[style]` bindings respectively.
PR Close#22719
This patch removes the deprecated support for animation
symbol imports from @angular/core.
BREAKING CHANGE: it is no longer possible to import
animation-related functions from @angular/core. All
animation symbols must now be imported from @angular/animations.
PR Close#22692
This change makes the code cleaner for the user. It does mean
a little bit more work for us since we have to patch the `type` back
into the `DirectiveDef`. However since the patching happens only once
on startup it should not be significant.
PR Close#21374
This separation is no longer needed since directives are now passed into the `container` as an array rather than as child functions of the `containerStart`
PR Close#21374
We used to have a separate `directive` instruction for instantiating
directives. However, such an instruction requires that directives
are created in the correct order, which would require that template
compiler would have knowledge of all dependent directives. This
would break template compilation locality principle.
This change only changes the APIs to expected form but does
not change the semantics. The semantics will need to be corrected
in subsequent commits. The semantic change needed is to
resolve the directive instantiation error at runtime based on
injection dependencies.
PR Close#21374
Each node now has two index: nodeIndex and checkIndex.
nodeIndex is the index in both the view definition and the view data.
checkIndex is the index in in the update function (update directives and update
renderer).
While nodeIndex and checkIndex have the same value for now, having both of them
will allow changing the structure of view definition after compilation (ie for
runtime translations).
With this commit `ngc` is used instead of `tsc-wrapped` for
collecting metadata and tsickle rewriting and `tsc-wrapped`
is removed from the repository.
`@angular/tsc-wrapped@5` is now deprecated and is no longer
used, updated, or maintained as part as of Angular 5.x.x.
`@angular/tsc-wrapped@4` is still maintained and required by
Angular 4.x.x and will be maintained as long as 4.x.x is in
LTS.
PR Close#19298
This change allows ReflectiveInjector to be tree shaken resulting
in not needed Reflect polyfil and smaller bundles.
Code savings for HelloWorld using Closure:
Reflective: bundle.js: 105,864(34,190 gzip)
Static: bundle.js: 154,889(33,555 gzip)
645( 2%)
BREAKING CHANGE:
`platformXXXX()` no longer accepts providers which depend on reflection.
Specifically the method signature when from `Provider[]` to
`StaticProvider[]`.
Example:
Before:
```
[
MyClass,
{provide: ClassA, useClass: SubClassA}
]
```
After:
```
[
{provide: MyClass, deps: [Dep1,...]},
{provide: ClassA, useClass: SubClassA, deps: [Dep1,...]}
]
```
NOTE: This only applies to platform creation and providers for the JIT
compiler. It does not apply to `@Compotent` or `@NgModule` provides
declarations.
Benchpress note: Previously Benchpress also supported reflective
provides, which now require static providers.
DEPRECATION:
- `ReflectiveInjector` is now deprecated as it will be remove. Use
`Injector.create` as a replacement.
closes#18496
- prevents unsubscribing from the zone on error
- prevents unsubscribing from directive `EventEmitter`s on error
- prevents detaching views in dev mode if there on error
- ensures that `ngOnInit` is only called 1x (also in prod mode)
Fixes#9531Fixes#2413Fixes#15925
The Router use the type `Params` for all of:
- position parameters,
- matrix parameters,
- query parameters.
`Params` is defined as follow `type Params = {[key: string]: any}`
Because parameters can either have single or multiple values, the type should
actually be `type Params = {[key: string]: string | string[]}`.
The client code often assumes that parameters have single values, as in the
following exemple:
```
class MyComponent {
sessionId: Observable<string>;
constructor(private route: ActivatedRoute) {}
ngOnInit() {
this.sessionId = this.route
.queryParams
.map(params => params['session_id'] || 'None');
}
}
```
The problem here is that `params['session_id']` could be `string` or `string[]`
but the error is not caught at build time because of the `any` type.
Fixing the type as describe above would break the build because `sessionId`
would becomes an `Observable<string | string[]>`.
However the client code knows if it expects a single or multiple values. By
using the new `ParamMap` interface the user code can decide when it needs a
single value (calling `ParamMap.get(): string`) or multiple values (calling
`ParamMap.getAll(): string[]`).
The above exemple should be rewritten as:
```
class MyComponent {
sessionId: Observable<string>;
constructor(private route: ActivatedRoute) {}
ngOnInit() {
this.sessionId = this.route
.queryParamMap
.map(paramMap => paramMap.get('session_id') || 'None');
}
}
```
Added APIs:
- `interface ParamMap`,
- `ActivatedRoute.paramMap: ParamMap`,
- `ActivatedRoute.queryParamMap: ParamMap`,
- `ActivatedRouteSnapshot.paramMap: ParamMap`,
- `ActivatedRouteSnapshot.queryParamMap: ParamMap`,
- `UrlSegment.parameterMap: ParamMap`
E.g. for a component like this:
```
@Component({
template: ‘<ng-content select=“child”></ng-content>’
})
class MyComp {
@Input(‘aInputName’)
aInputProp: string;
@Output(‘aEventName’)
aOuputProp: EventEmitter<any>;
}
```
the `ComponentFactory` will now contain the following:
- `inputs = {aInputProp: ‘aInputName’}`
- `outputs = {aOutputProp: ‘aOutputName’}`
- `ngContentSelectors = [‘child’]`
E.g. for a component like this:
```
@Component({
template: ‘<ng-content select=“child”></ng-content>’
})
class MyComp {
@Input(‘aInputName’)
aInputProp: string;
@Output(‘aEventName’)
aOuputProp: EventEmitter<any>;
}
```
the `ComponentFactory` will now contain the following:
- `inputs = {aInputProp: ‘aInputName’}`
- `outputs = {aOutputProp: ‘aOutputName’}`
- `ngContentSelectors = [‘child’]`
fixes#12869fixes#12889fixes#13885fixes#13870
Before this change there was a single injector tree.
Now we have 2 injector trees, one for the modules and one for the components.
This fixes lazy loading modules.
See the design docs for details:
https://docs.google.com/document/d/1OEUIwc-s69l1o97K0wBd_-Lth5BBxir1KuCRWklTlI4
BREAKING CHANGES
`ComponentFactory.create()` takes an extra optional `NgModuleRef` parameter.
No change should be required in user code as the correct module will be used
when none is provided
DEPRECATIONS
The following methods were used internally and are no more required:
- `RouterOutlet.locationFactoryResolver`
- `RouterOutlet.locationInjector`
DEPRECATION:
Use `RouterModule.forRoot(routes, {initialNavigation: 'enabled'})` instead of
`RouterModule.forRoot(routes, {initialNavigtaion: true})`.
Before doing this, move the initialization logic affecting the router
from the bootstrapped component to the boostrapped module.
Similarly, use `RouterModule.forRoot(routes, {initialNavigation: 'disabled'})`
instead of `RouterModule.forRoot(routes, {initialNavigation: false})`.
Deprecated options: 'legacy_enabled', `true` (same as 'legacy_enabled'),
'legacy_disabled', `false` (same as 'legacy_disabled').
The "Router Initial Navigation" design document covers this change.
Read more here:
https://docs.google.com/document/d/1Hlw1fPaVs-PCj5KPeJRKhrQGAvFOxdvTlwAcnZosu5A/edit?usp=sharing
Previously, the relative order of the AngularJS compiling/linking operations was
not similar to AngularJS's, resulting in inconsistent behavior for upgraded
components (which made upgrading to Angular less straight forward).
This commit fixes it, by following the compiling/linking process of AngularJS
more closely.
Main differences:
- The components view is already populated when the controller is instantiated
(and subsequent hooks are called).
- The correct DOM content is available when running the `$onChanges`, `$onInit`,
`$doCheck` hooks. Previously, the "content children" were still present, not
the "view children".
- The same for pre-linking.
- The template is compiled in the correct DOM context (e.g. has access to
ancestors). Previously, it was compiled in isolation, inside a dummy element.
For reference, here is the order of operations:
**Before**
1. Compile template
2. Instantiate controller
3. Hook: $onChanges
4. Hook: $onInit
5. Hook: $doCheck
6. Pre-linking
7. Collect content children
8. Insert compiled template
9. Linking
10. Post-linking
11. Hook: $postLink
**After**
1. Collect content children
2. Insert template
3. Compile template
4. Instantiate controller
5. Hook: $onChanges
6. Hook: $onInit
7. Hook: $doCheck
8. Pre-linking
9. Linking
10. Post-linking
11. Hook: $postLink
Fixes#13912
Secondary entry points (testing, static, etc) are rolled up into a
single ESM/ES2015 file, then downleveled to ESM/ES2015. This downleveling
was not working and was producing ESM/ES2015. Also, the @angular/core
package's .babelrc file was missing reference to Rx Observable which
broke the UMD bundle.
Fixes#14730
Miško Hevery
Ben Lesh
Matias Niemelä
Alex Eagle
Adam Plumer
Greg Magolan
Alex Rickabaugh
Peter Johan Salomonsen
Wassim Chegham
Paul Gschwendtner
Kara Erickson
Marc Laval
Daniel Ruf
Igor Minar
George Kalpakas
Rob Wormald
Veres Lajos
Victor Berchet
David-Emmanuel Divernois
Alex Eagle
Olivier Combe
Chuck Jazdzewski
Tobias Bosch
Jason Aden
Victor Savkin
Dzmitry Shylovich
Igor Minar
Igor Minar
Jason Aden
vikerman