Before this commit we were creating a "fake" TNode for each and every
projectable node passed during dynamic component creation. This approach
had several problems:
- the existing TView structure had to be mutated to accomodate new TNodes and
it was very easy to "corrupt" TView / TNode data structures;
- TNodes are not really needed to fully support projectable nodes so we were
creating objects and updating existing data structures for nothing.
This commit changes the approach so we don't create "fake" TNodes for projectable
nodes but instead we process projectable nodes directly in the projection instruction.
As a result we've got less code, less object allocation and - as a bonus - we fix few
bugs where TView / TNode data structures were corrupted when using projectable nodes.
PR Close#28275
Fixes Ivy not applying properties that are set in camelCase, because it goes through the `CSSStyleDeclaration` API via `setProperty` and `removeProperty` which requires for [the values to be in dash-case](https://developer.mozilla.org/en-US/docs/Web/API/CSSStyleDeclaration/setProperty).
**Note:** I opted to let the browser normalize the value, rather than convert it to dash-case during compile time, because there are some special cases like browser-prefixed properties where we might not normalize it in-line with the browser.
This PR fixes FW-579.
PR Close#28276
Previous to this change, the isStylingContext() function was improperly
returning true for LContainers because it used the presence of an array
at index 2 to determine whether it was a styling context. Unfortunately,
LContainers also contain arrays at index 2, so this would return a false
positive. This led to other errors down the line because we would treat
nodes with containers as if they already had styling contexts (even if
they did not), so the proper initialization logic for styling contexts
was not run.
This commit fixes the isStylingContext() function to use LCONTAINER_LENGTH
as a marker rather than the presence of an array, which in turn fixes
host bindings to styles on nodes with containers.
PR Close#28221
With the refactoring or how styles/classes are implmented in Ivy,
interpolation has caused the binding code to mess up since interpolation
itself takes up its own slot in Ivy's memory management code. This patch
makes sure that interpolation works as expected with class and style
bindings.
Jira issue: FW-944
PR Close#28190
In VE the renderer.begin() and renderer.end() methods are only called
when CD is called on an element. This patch ensures that Ivy does the
same thing.
Jira issue: FW-945
PR Close#28192
There were two issues with multiple ICU expressions in the same i18n block:
- the regexp that was used to parse the text wasn't able to handle multiple ICU expressions, I've replaced it with parsing the text and searching for brackets (which is what we ended up doing in the end anyway)
- we allocate node indexes for nodes generated by the ICU expressions which increases the expando value, but we would create the nodes for those cases during the update phase. In the mean time we would create some nodes during the creation phase (comment nodes for ICU expressions, text nodes, ...) with an auto increment index. This means that any node created after an ICU expression would get the following index value, but the ICU case nodes expected to use the same index as well... There was a mismatch between the auto generated index, and the expected index which was causing problems when we needed to select those nodes for updates later on. To fix it, I've added the expected node index to the list of mutate codes that we generate, and we do not use an auto increment value anymore.
FW-905 #resolve
PR Close#28083
Up until this point, all static attribute values (things like `title` and `id`)
defined within the `host` are of a Component/Directive definition were
generated into a `def.attributes` array and then processed at runtime.
This design decision does not lend itself well to tree-shaking and is
inconsistent with other static values such as styles and classes.
This fix ensures that all static attribute values (attributes, classes,
and styles) that exist within a host definition for components and
directives are all assigned via the `elementHostAttrs` instruction.
```
// before
defineDirective({
...
attributes: ['title', 'my title']
...
})
//now
defineDirective({
...
hostBindings: function() {
if (create) {
elementHostAttrs(..., ['title', 'my-title']);
}
...
}
...
})
```
PR Close#28089
Angular allows for `<ng-content>` elements to include a selector which
filters which content-projected entries are inserted into the container
depending on whether or not the selector is matched.
With Ivy this feature has not fully worked due to the massive changes
that took place inside of Ivy's styling algorithm code (which is
responsible for assigning classes and styles to an element). This
fix ensures that content-projection can correctly identify which slot
an element should be placed into when class-based selectors are used.
PR Close#27849
When requesting a queries instance for a node, it was previously
decided whether it needs to be cloned if the node was not already marked
as hosting a query. This check is in place to have only a single queries
instance per node.
The issue with this approach is that no clone is created for subsequent
instantiations of a component, as the TNode is already marked as hosting
a query during first template pass, whereas the cloning of queries
should be independent of first template pass.
To overcome this issue, the queries are assigned an owner TNode such
that it can reliably be determined if a clone needs to be created.
PR Close#27892
This change is a prerequasity for a later change which will turn the
'di' into its own bazel package. In order to do that we have to:
- have `Injector` type be importable by Ivy. This means that we need
to create `Injector` as a pure type in `interface` folder which is
already a bazel package which Ivy can depend on.
- Remove the dependency of `class Injector` on Ivy so that it can be
compiled in isolation. We do that by using `-1` as special value for
`__NG_ELEMENT_ID__` which tells the Ivy `NodeInjector` than
`Injector` is being requested.
PR Close#28066
Previously, we had the logic to schedule a change detection tick
inside markViewDirty(). This is fine when used in markDirty(),
the user-facing API, because it should always schedule change
detection. However, this doesn't work when used in markForCheck()
because historically markForCheck() does not trigger change
detection.
To be backwards compatible, this commit moves the scheduling
logic out of markViewDirty() and into markDirty(), so
markForCheck no longer triggers a tick.
PR Close#28048
Libraries that create components dynamically using component factories,
such as `@angular/upgrade` need to pass blocks of projected content
through to the `ComponentFactory.create()` method. These blocks
are extracted from the content by matching CSS selectors defined in
`<ng-content select="..">` tags found in the component's template.
The Angular compiler collects these CSS selectors when compiling a component's
template, and exposes them via the `ComponentFactory.ngContentSelectors`
property.
This change ensures that this property is filled correctly when the
component factory is created by compiling a component with the Ivy engine.
PR Close#27867
This commit adds sanitization for `elementProperty` and `elementAttribute` instructions used in `hostBindings` function, similar to what we already have in the `template` function. Main difference is the fact that for some attributes (like "href" and "src") we can't define which SecurityContext they belong to (URL vs RESOURCE_URL) in Compiler, since information in Directive selector may not be enough to calculate it. In order to resolve the problem, Compiler injects slightly different sanitization function which detects proper Security Context at runtime.
PR Close#27939
This update introduces support for global object (window, document, body) listeners, that can be defined via host listeners on Components and Directives.
PR Close#27772
`i18nAttributes` was throwing an error when it was called multiple times in the create part of the template function with the same index, for example when we create multiple components with the same template. It shouldn't throw in this case, and just use the cache when available.
FW-903 #resolve
PR Close#27911
Fixes Ivy's `QueryList` not being an instance of the exported ViewEnginer `QueryList`.
Also reworks `first`, `last` and `length` to be regular properties, rather than setters. Reworking `length` was required to be able to extend the ViewEngine `QueryList`, but I reworked `first` and `last` as well since getters generate a lot more code when transpiled to ES5.
These changes fix FW-706.
PR Close#27942
This PR assures that content projection works if an <ng-content> tag is
placed inside an <ng-template> in one component and that <ng-template>
is inserted into a different component. It fixes a bug where the
projection instruction code would walk up the insertion tree to find
selector data instead of the declaration tree.
PR Close#27783
Prior to this change, ICU extraction logic was not taking into account nested bindings (that look like this: �0:1�) and only accounted for top level bindings (like this �0�). As a result, ICUs were not parsed and remained as text in the output. Now the extraction logic (regular expressions) take into account the nested bindings format as well.
PR Close#27914
Prior to this commit, we had two different modes for change detection
execution for Ivy, depending on whether you called `bootstrap()` or
`renderComponent()`. In the former case, we would complete creation
mode for all components in the tree before beginning update mode for
any component. In the latter case, we would run creation mode and
update mode together for each component individually.
Maintaining code to support these two different execution orders was
unnecessarily complex, so this commit aligns the two bootstrapping
mechanisms to execute in the same order. Now creation mode always
runs for all components before update mode begins.
This change also simplifies our rendering logic so that we use
`LView` flags as the source of truth for rendering mode instead of
`rf` function arguments. This fixed some related bugs (e.g. calling
`ViewRef.detectChanges` synchronously after the view's creation
would create view nodes twice, view queries would execute twice, etc).
PR Close#27744
We invoked `hostBindings` function in Create and Update modes with different element index due to the fact that we did not subtract HEADER_OFFSET from the index before passing it to `hostBindings` function in Create mode. Now we subtract HEADER_OFFSET value before invoking `hostBindings`, which makes Ceate and Update calls consistent.
PR Close#27694
Context discovery was only available on elements. This PR adds support for containers and ICU expressions.
FW-378 #resolve
FW-665 #comment linker integration tests
PR Close#27644
`NgModule` requires that `Component`s/`Directive`s/`Pipe`s are listed in
declarations, and that each `Component`s/`Directive`s/`Pipe` is declared
in exactly one `NgModule`. This change adds runtime checks to ensure
that these sementics are true at runtime.
There will need to be seperate set of checks for the AoT path of the
codebase to verify that same set of semantics hold. Due to current
design there does not seem to be an easy way to share the two checks
because JIT deal with references where as AoT deals with AST nodes.
PR Close#27604
Prior to this change, we were unable to match directives using `ng-template` tags (for example the following selector would not work even though there might be some <ng-template>s in a template: `ng-template[directiveA]`. As a result, that broke some components that relies on such selectors to work. In order to resolve the problem, we now pass tag name to the `template` instruction (where we passed `null` before) and this tag name is used for matching at runtime. This update should also help support projecting containers, because the tag name is required to properly match such elements.
PR Close#27636
In Ivy, a pure call to `setClassMetadata` is inserted to retain the
information that would otherwise be lost while eliding the Angular
decorators. In the past, the Angular constructor decorators were
wrapped inside of an anonymous function which was only evaluated once
`ReflectionCapabilities` was requested for such metadata. This approach
prevents forward references from inside the constructor parameter
decorators from being evaluated before they are available.
In the `setClassMetadata` call, the constructor parameters were not wrapped
within an anonymous function, such that forward references were evaluated
too early, causing runtime errors.
This commit changes the `setClassMetadata` call to pass the constructor
parameter decorators inside of an anonymous function again, such that
forward references are not resolved until requested by
`ReflectionCapabilities`, therefore avoiding the early reads of forward refs.
PR Close#27561
Currently the `ViewRef.destroy` method assumes that its index inside the view container will always be valid, however if it has been removed already, it'll be -1 which will throw an error.
The error manifested itself in one of the unit tests where a view had been detached during the test and then `TestBed` attempted to destroy its `ComponentRef` which ended threw an `Error during cleanup of component`.
PR Close#27585
Since Renderer is shared across root and child views, we need to avoid `destroy` method invocation for child views and only invoke is for root view when needed. Prior to this change, the `destroy` function was called whenever child view was destroyed, thus causing errors at runtime.
PR Close#27592
Previously in Ivy, host bindings did not work if they shared a public name
with an Input because they used the `elementProperty` instruction as is.
This instruction was originally built for inside component templates, so it
would either set a directive input OR a native property. This is the
correct behavior for inside a template, but for host bindings, we always
want the native properties to be set regardless of the presence of an Input.
This change adds an extra argument to `elementProperty` so we can tell it to
ignore directive inputs and only set native properties (if it is in the
context of a host binding).
PR Close#27589
In `ViewRef.detectChanges`, we are passing `ViewRef.context` into `detectChanges` to trigger change detection. This only makes sense for component `ViewRefs` (i.e. injected `ChangeDetectorRefs`) because with embedded views, `context` is not a component instance where the view has been monkey-patched. It's a just a normal object, so the view will be undefined.
In order to resolve this problem, we now invoke `detectChangesInternal` and also pass `LView` (to make sure we always have a view available).
PR Close#27521
Prior to this change, animation properties were defined as element attributes, which caused errors at runtime. Now all animation-related attributes are defined as element properties.
Also as a part of this update, we start to account for bindings used in animations, which was previously missing.
PR Close#27496
Prior to this change, the number of host vars stored for directives with `hostBindings` in expando block was incorrect for inherited directives (in case both parent and child directive have `hostBindings` defined). Now if we identify that we already added a `hostBinding` into expando block, we just increase the corresponding number of host binding vars
PR Close#27392
In Angular, it used to be an accepted practice to use strings as dependency
injection tokens. E.g. {provide: 'test', useValue: 'provided'}. However,
the Ivy node injection system did not support this. The Ivy DI system
attempts to patch a Bloom bit index onto each type registered with it, and
this patch operation does not work for a string token.
This commit adds string token support to the bloom filter system by
reserving bit 0 for string tokens. This eliminates the need for each string
token to store its own Bloom bit, at the expense of slightly more expensive
lookups of string tokens.
PR Close#27383
BREAKING CHANGE:
The public API for `DebugNode` was accidentally too broad. This change removes
1. Public constructor. Since `DebugNode` is a way for Angular to communicate information
on to the developer there is no reason why the developer should ever need to
Instantiate the `DebugNode`
2. We are also removing `removeChild`, `addChild`, `insertBefore`, and `insertChildAfter`.
All of these methods are used by Angular to constructor the correct `DebugNode` tree.
There is no reason why the developer should ever be constructing a `DebugNode` tree
And these methods should have never been made public.
3. All properties have been change to `readonly` since `DebugNode` is used by Angular
to communicate to developer and there is no reason why these APIs should be writable.
While technically breaking change we don’t expect anyone to be effected by this change.
PR Close#27223
This API is part of our public api surface and needs to be monitored by the public_api_guard.
I also had to go back and mark all of the exported functions with @publicApi jsdoc tag.
PR Close#27008
This commit introduces the setClassMetadata() private function, which
adds metadata to a type in a way that can be accessed via Angular's
ReflectionCapabilities. Currently, it writes to static fields as if
the metadata being added was downleveled from decorators by tsickle.
The plan is for ngtsc to emit code which calls this function, passing
metadata on to the runtime for testing purposes. Calls to this function
would then be tree-shaken away for production bundles.
Testing strategy: proper operation of this function will be an integral
part of TestBed metadata overriding. Angular core tests will fail if this
is broken.
PR Close#26860
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
We are close enough to blacklist a few test targets, rather than whitelist targets to run...
Because bazel rules can be composed of other rules that don't inherit tags automatically,
I had to explicitly mark all of our ts_library and ng_module targes with "ivy-local" and
"ivy-jit" tags so that we can create a query that excludes all fixme- tagged targets even
if those targets are composed of other targets that don't inherit this tag.
This is the updated overview of ivy related bazel tags:
- ivy-only: target that builds or runs only under ivy
- fixme-ivy-jit: target that doesn't yet build or run under ivy with --compile=jit
- fixme-ivy-local: target that doesn't yet build or run under ivy with --compile=local
- no-ivy-jit: target that is not intended to build or run under ivy with --compile=jit
- no-ivy-local: target that is not intended to build or run under ivy with --compile=local
PR Close#26471
Originally, the ivy_switch mechanism used Bazel genrules to conditionally
compile one TS file or another depending on whether ngc or ngtsc was the
selected compiler. This was done because we wanted to avoid importing
certain modules (and thus pulling them into the build) if Ivy was on or
off. This mechanism had a major drawback: ivy_switch became a bottleneck
in the import graph, as it both imports from many places in the codebase
and is imported by many modules in the codebase. This frequently resulted
in cyclic imports which caused issues both with TS and Closure compilation.
It turns out ngcc needs both code paths in the bundle to perform the switch
during its operation anyway, so import switching was later abandoned. This
means that there's no real reason why the ivy_switch mechanism needed to
operate at the Bazel level, and for the ivy_switch file to be a bottleneck.
This commit removes the Bazel-level ivy_switch mechanism, and introduces
an additional TypeScript transform in ngtsc (and the pass-through tsc
compiler used for testing JIT) to perform the same operation that ngcc
does, and flip the switch during ngtsc compilation. This allows the
ivy_switch file to be removed, and the individual switches to be located
directly next to their consumers in the codebase, greatly mitigating the
circular import issues and making the mechanism much easier to use.
As part of this commit, the tag for marking switched variables was changed
from __PRE_NGCC__ to __PRE_R3__, since it's no longer just ngcc which
flips these tags. Most variables were renamed from R3_* to SWITCH_* as well,
since they're referenced mostly in render2 code.
Test strategy: existing test coverage is more than sufficient - if this
didn't work correctly it would break the hello world and todo apps.
PR Close#26550
This commit adds an ngTemplateGuard_ngIf static method to the NgIf
directive and an ngTemplateContextGuard static method to NgFor. The
function of these two static methods is to enable type narrowing
within generated type checking code for consumers of the directives.
PR Close#26203
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
Various user code uses 'instanceof' to check whether a particular instance
is a TemplateRef, ElementRef, etc. Ivy needs to work with these checks.
PR Close#25775
defineComponent() and friends can return a flyweight EMPTY object for
specific fields when they contain no data. InheritDefinitionFeature
was attempting to write into these flyweight objects, which have been
protected with Object.freeze().
This commit adds detection to InheritDefinitionFeature to identify the
frozen objects.
PR Close#25755
This commit creates an API for factory functions which allows them
to be inherited from one another. To do so, it differentiates between
the factory function as a wrapper for a constructor and the factory
function in ngInjectableDefs which is determined by a default
provider.
The new form is:
factory: (t?) => new (t || SomeType)(inject(Dep1), inject(Dep2))
The 't' parameter allows for constructor inheritance. A subclass with
no declared constructor inherits its constructor from the superclass.
With the 't' parameter, a subclass can call the superclass' factory
function and use it to create an instance of the subclass.
For @Injectables with configured providers, the factory function is
of the form:
factory: (t?) => t ? constructorInject(t) : provider();
where constructorInject(t) creates an instance of 't' using the
naturally declared constructor of the type, and where provider()
creates an instance of the base type using the special declared
provider on @Injectable.
PR Close#25392
Ben Lesh
Pawel Kozlowski
Kristiyan Kostadinov
Kara Erickson
Matias Niemelä
Olivier Combe
Marc Laval
JoostK
Misko Hevery
Pete Bacon Darwin
Andrew Kushnir
Alex Eagle
George Kalpakas
Alex Rickabaugh
Igor Minar
Greg Magolan
Adam Plumer
Victor Berchet