Reworks the pure functions to use arrow functions with an implicit return instead of function expressions. This allows us to shave off some bytes for each pure function, because we can avoid some of the syntax.
PR Close#51668
These changes build on top of #51514 to add support for advanced expressions inside the `track` parameter of `for` loop blocks. There are two different outputs that the compiler can generate:
1. If the tracking function only references the item or `$index`, the compiler generates a pure arrow function as a constant references in the `repeaterCreate` instruction.
2. If the tracking function has references to properties outside of the `for` loop block, the compiler will rewrite those references to go through `this` and generate a function declaration. The runtime will `bind` the declaration to the current component instance so that the rewritten `this` references are resolved correctly.
Advanced tracking expression come with the following limitations to ensure the best possible performance:
1. They can only reference the item, `$index` and properties directly on the component instance. This means that there'll be an error when accessing this like local template variables and references. While we could get this to work, we would have to traverse the context tree at runtime which will degrade the performance of the loop, because it's a linear time operation that is performed on each comparison. Furthermore, allowing local references would require us re-evaluate the list when any one of them has changed.
2. Pipes aren't allowed inside the tracking function.
3. Object literals and pipes used inside the tracking expression will be recreated on each invocation.
PR Close#51618
Adds type checking for the contents of `if`, `switch` and `for` blocks.
**Note:** this is just an initial implementation to get some basic type checking working and to figure out the testing setup. We'll need special TCB structures for this syntax so that we can support type narrowing.
PR Close#51570
With the new control flow and defer blocks it'll be common for several template instructions to be declare one after another. These changes add support for chaining to the `template` instruction which will allow us to save some bytes.
PR Close#51546
Adds the initial implementation to generate the instructions for the `for` loop block.
**Note:** the expressions we support in the `track` paramateter are currently limited to tracking by identity or index, or a specific property of the item. Supporting more advanced expression will require additional work that I'll do in a follow-up PR.
PR Close#51514
`switch` blocks are part of the new control flow syntax. This commit adds support for processing them, and emitting the appropriate templates and conditional instructions.
PR Close#51518
`syntheticHostListener` and `listener` have ordering dependencies. We reuse the existing ordering phase, and generalize it to also order create mode instructions.
PR Close#51498
Animation listeners on host bindings result in a special `syntheticHostListener` instruction. We can now emit this instruction.
Additionally, the naming phase for events has been slightly refactored to smoothly incorporate whether the event is from a host listener, as well as whether it is an animation listener.
PR Close#51498
For host bindings, `TemplateDefinitionBuilder` seems to use a different binding ordering, in which style bindings come after all the property bindings. We approximate that by treating `hostProperty` differently from `property` in the ordering phase.
PR Close#51498
The template pipeline is already capable of parsing and processing class and style attributes on templates. We now extend that functionality to host bindings.
The parser, for some reason, splits out class and style attributes into a `specialAttributes` field. We merge them back into the main attributes map, and allow the template pipeline to process them normally.
PR Close#51498
TemplateDefinitionBuilder only extracts host attributes if they are text attributes. For example, `[attr.foo]="'my-value'"` is not extracted despite being a string literal, because it is not a text attribute.
PR Close#51498
Host property bindings can be animation bindings, and should be ingested and emitted as such, as well as being processed by the renaming phase.
PR Close#51498
Host bindings can apply static attributes. These will be extracted to a `hostAttrs` field on the host binding function's metadata.
In order to achieve this, we add an `attributes` field to the host binding job. Then, we peform attribute exraction on host bindings. We finally populate the `attributes` field directly, instead of relying on a `consts` array.
PR Close#51498
Adds i18n block start & end ops, as well as a new phase to construct the
i18n message variable to be added to the consts array.
Co-authored-by: Alex Rickabaugh <alx+alxhub@alxandria.net>
Co-authored-by: Dylan Hunn <dylhunn@gmail.com>
PR Close#51353
This commit adds an initial implementation of the `{#defer}` block runtime, which supports the `when` conditions. More conditions and basic prefetching support will be added in followup PRs.
PR Close#51347
Adds the logic to generate the instructions for `if` blocks. There are two primary use cases we need to account for:
A conditional that doesn't use the `as` parameter of the `if` block. To support it we generate a nested ternary expression that evaluates to the index of the template whose condition is truthy. If the block doesn't have an `else` branch, we pass in a special `-1` value which means that no view will be rendered.
Example with an `else`:
```ts
// {#if expr}
// ...
// {:else if otherExpr} ...
// {:else} ...
// {/if}
if (rf & 1) {
ɵɵtemplate(0, App_Conditional_0_Template, 0, 0);
ɵɵtemplate(1, App_Conditional_1_Template, 0, 0);
ɵɵtemplate(2, App_Conditional_2_Template, 0, 0);
}
if (rf & 2) {
ɵɵconditional(0, ctx.expr ? 0 : ctx.otherExpr ? 1 : 2);
}
```
Example without an `else`:
```ts
// {#if expr}
// ...
// {:else if otherExpr} ...
// {/if}
if (rf & 1) {
ɵɵtemplate(0, App_Conditional_0_Template, 0, 0);
ɵɵtemplate(1, App_Conditional_1_Template, 0, 0);
}
if (rf & 2) {
ɵɵconditional(0, ctx.expr ? 0 : ctx.otherExpr ? 1 : -1);
}
```
If a conditional captures it's value in an alias (e.g. `{#if expr; as foo}`) we need to assign the value to a temporary variable before passing it along to `conditional`.
```ts
// {#if expr; as alias}...{/if}
if (rf & 1) {
ɵɵtemplate(0, App_Conditional_0_Template, 1, 0);
}
if (rf & 2) {
let App_contFlowTmp;
ɵɵconditional(0, (App_contFlowTmp = ctx.expr) ? 0 : -1, App_contFlowTmp);
}
```
PR Close#51380
Angular 16.1 introduced the input transform feature, requiring the partial compilation output to be extended
with a reference to the input transform function. This has resulted in a subtle breaking change, where older
versions of the Angular linker can no longer consume libraries that have started to use this feature.
We do try to support using a 16.1 library from an Angular 16.0 application, but if a library actually
adopts a new feature then this is no longer possible. In such cases, it is desirable to report a message
telling the user that their version of the Angular compiler is too old, as determined by the `"minVersion"`
property that is present in each partial declaration. This version would still indicate that the declaration
required at least Angular 14.0 to be compiled, but this is not accurate once input transforms are being
used. Consequently, this error would not be reported, causing a less informative error once the input transform
was being observed.
Fixes#51411
PR Close#51413
In local compilation mode it is not possible to use an imported string for component's template or styles as it cannot be resolved statically in compile time. There are some such use cases in g3 and potentially devs might incorporate such pattern. At the moment such pattern will cause the local compilation fail with generic error messages (e.g., so and so at position 1 is not a reference, etc). This change makes specific error messages with helpful hints for such cases. These new error messages can help devs to quickly resolve the issue as well as make it possible to identify existing issues in g3.
PR Close#51338
`tsickle` is not used in any code paths in 3P and we can remove
this complexity. The `tsickle` npm package has not been released
in a while and we are risking breakages with e.g. future TypeScript
versions.
Note that the `ng_module` rule was updated to not emit through
tsickle at all. The tsickle in 1P is done directly by `tsc_wrapped`
and our code path in `compiler-cli` is not needed at all.
PR Close#50602
This commit updates TestBed to wait for async component metadata resolution before compiling components.
Async metadata is added by the compiler in case a component uses defer blocks, which contain deferrable
symbols.
PR Close#51182
This commit updates compiler logic to generate the `setClassMetadataAsync` calls for components that used defer blocks. The `setClassMetadataAsync` function loads deferrable dependencies and invokes the `setClassMetadata` synchronously once everything is loaded. This change is needed to avoid eager references to deferrable symbols in component metadata in generated code.
PR Close#51182
Fixes that we weren't processing `when` conditions correctly which led to a compilation error when a pipe is used inside the expression.
PR Close#51368
A factory generator function called "i0.ɵɵgetComponentDepsFactory" is added to generate a factory function for component dependencies. This function will use the deps tracker to calculate the component's dependencies.
For standalone components the component imports (if exists) will be passed to this function. Alternatively this function can grab the imports directly from the decorate, but such extraaction needs some runtime logic which overlapps with what the trait compiler is doing. So better to pass the imports directly to this function at compile time.
PR Close#51089
In local mode the compiler combines the raw imports and exports and pass them to the injector definition as the imports field. It is not possible to filter out ng modules at compile time though, and it will be done in runtime.
Unit tests also added, and since that was the first time adding tests for local compilation some tweaks had to be made in order to disable diagnostics in local compilation mode in order for tests to run (such situation is also the case in real compilation where we ignore all teh diagnostics basically)
PR Close#51089
This commit updates the logic to drop regular imports when all symbols that it brings can be defer-loaded.
The change ensures that there is no mix of regular and dynamic imports present in a source file.
PR Close#51171
Updates the template pipeline's temporary variables phase to reuse
temporary variables within an expression. The algorithm implemented here
reuses variables more aggressively than TemplateDefinitionBuilder. This
change in behavior is acceptable, as it is unlikely to cause any
failures, and implementing the exact behavior observed in
TemplateDefinitionBuilder would be difficult.
PR Close#51100
In some cases it is not feasible to have the template pipeline produce
the exact same compiled output as the TemplateDefinitionBuilder. This
commit adds support to the testing infrastructure to have different
expected output files for each. This option should be used sparingly, as
we want the output to be as close as possible.
PR Close#51100
Updates the TemplateDefinitionBuilder class to generate the `defer` instruction for `{#defer}` blocks. Also generates dependency function that would be invoked at runtime (with dynamic imports inside).
PR Close#51162
Host property bindings beginning with `attr.` should have `Attribute` binding kind, and result in an `attribute` instruction.
This should really be handled in the parser in the future.
PR Close#51188
Templates may contain special `svg` and `math` elements, as well as logical descendants of those elements (e.g. `svg` may contain `g`). These will be parsed with a special colon-prefixed *namespace identifier*, such as `:svg:svg`, or `:svg:g`, or `:math:infinity`.
The template pipeline now considers these namespace prefixes, and stores them specially on the Element and Template data structures, ultimately generating the appropriate runtime instructions to change namespaces when needed.
PR Close#51188
When a container-like element has the `ngNonBindable` special attribute, bindings are disabled for it and its descendants. This requires emitting the `disableBindings` and `enableBindings` instructions when nested content exists.
PR Close#51188
Interestingly, host bindings are parsed quite differently from template functions. For example, bindings such as `[style.foo]: 3px` would be parsed into a value, unit, and type when bound to a template, but will not be parsed as such when used in a host binding.
In this commit, we remedy this shortcoming by adding support for bindings in host binding functions to the template pipeline. In particular, we create a phase to process these bindings, and transform them into the correct output binding kind.
Additionally, we fix some other minor bugs and omissions.
Finally, we enable compilation of host bindings with the template pipeline, which requires us to turn off a number of failing tests.
PR Close#50899
Begin producing source maps for the template pipeline, for a couple fundamental kinds of instructions, including elements, templates, properties, text, and interpolations.
PR Close#50899
Previously, `$event` was interpreted as a lexical read on the enclosing context. Now, a new pass converts such reads into simple output AST reads of `$event`, so they are not processed by the context resolution or naming phases. Additionally, the same pass sets a field on the enclosing listener op, so that the reify phase does not have to search for reads of `$event`.
PR Close#50899
`$any(...)` casts should be dropped, except when they are an explicit call on `this.$any(...)`. Fix a bug in which we were transforming `ThisReceiver` into an implicit receiver.
PR Close#50899
Ensures that all property and attribute ops are ordered consistently
regardless of the order they appear in the template. This ensures
correct precedence (e.g. `[style.color]="'#000'"` awlays wins out over
`[style]="{color: '#fff'}"`)
PR Close#50805
An internal compiler option named `supportJitMode` is now available for use by the Angular CLI.
This option currently controls the emit of NgModule selector scope information. This emitted
information is only needed in AOT mode when an application also uses JIT. However, AOT mode
combined with JIT mode is not currently supported nor will work in the Angular CLI. With
the Angular CLI, JIT mode is only supported if the entire application is built in JIT mode.
Without this option, the CLI needs to manually perform a code transform to remove the information
and also replicate TypeScript's import eliding. This is can be a complicated operation and must
be continually kept up to date with any changes to both the Angular compiler and TypeScript.
The introduction of this new option alleviates these concerns while also removing several build
time actions that would otherwise need to be performed on every application build.
PR Close#51007
This commit adds the ability to generate attribute instructions as a result of property bindings such as `[attr.foo]='bar'` or `attr.foo='{{bar}}'`. "Singleton" interpolations, such as the previous example, will also be transformed into a simple `attribute` instruction.
PR Close#50818
The option 'local compile' is added for the test cases, and the locally compiled file for an input `abc.ts` is compared by default with the file `abc.local.js`. This allows to use the same input `abc.ts` for both full compilation (compared with `abc.js`) and local compilation (compared with `abc.local.js`). An example is provided in the next commit when compliance tests are added for the NgModule local compilation.
PR Close#50577
The expression `a()?.b` should expand into `(tmp = a()) === null ? null : tmp.b`, in order to avoid calling the function `a()` twice.
This commit modifies the null-safe-expansion algorithm to emit temporary assignments, and provides the reification code to actually generate the declarations, assignments, and reads.
Note also that, with our bottom-up algorithm, there are some tricky cases when a function call exists inside an indexed access, such as `f1()?.[f2()?.a]?.b`. We add some special logic to avoid generating a double-assignment to the temporary storing the result of `f2()`.
Finally, there are opportunities to reuse the same temporary in expressions like `a?.[f()]?.[f()]`. We save this for the next commit.
PR Close#50688
An internal compiler option named `supportTestBed` is now available for use by the
Angular CLI. This option currently controls the extraction and emit of Angular class
metadata. This emitted information is only needed in AOT mode when using certain
TestBed APIs. However, AOT mode is currently not available for unit testing within
the Angular CLI. As a result, the metadata is not used within CLI generation applications
and in particular production applications. Without this option, the CLI needs to
manually perform a code transform to remove the metadata and also replicate TypeScript's
import eliding. This is can be a complicated operation and must be continually kept
up to date with any changes to both the Angular compiler and TypeScript. The introduction
of this new option alleviates these concerns.
PR Close#50604
A larger investigation on why this is flaky is needed. Currently the
test is flaky with around 77% sucess rate and negatively impacts team
productivity. Subjectively, as reported by team members this it's much
more flaky than a success rate of 77%.
PR Close#50718
If a library is compiling with Angular v16.1.0, the library will break
for users that are still on Angular v16.0.x. This happens because the
`DirectiveDeclaration` or `ComponentDeclaration` types are not expecting
an extra field for `signals` metadata. This field was only added to the
generic types in `16.1.0`- so compilations fail with errors like this:
```
Error: node_modules/@angular/material/icon/index.d.ts:204:18 -
error TS2707: Generic type 'ɵɵComponentDeclaration' requires between 7 and 9 type arguments.
```
To fix this, we quickly roll back the code for inserting this metadata
field. That way, libraries remain compatible with all v16.x framework
versions.
We continue to include the `signals` metadata if `signals: true` is set.
This is not public API anyway right now- so cannot happen- but imagine
we expose some signal APIs in e.g. 16.2.x, then we'd need this metadata
and can reasonably expect signal-component library users to use a more
recent framework core version.
PR Close#50714
Angular's null-safe access operators differ from Javascript's built-in semantics, in that they short-circuit to `null` instead of `undefined`. This necessitates providing a custom transformation, instead of relying on Typescript or Javascript itseld.
The old TemplateDefinitionBuilder uses a top-down approach based on the Visitor pattern, in which it recursively extracts the left-most safe access, and hoists it to a null check at the top. See `expression_converter.ts` for details.
In this commit, we replace that approach with a new bottom-up algorithm, as part of the template pipeline. This requires an intermediate expression type to represent the not-yet-expanded ternary operators, and is split into its own pass.
Null-safe function calls are not yet implemented, since they will rely on a future temporary variable allocation pass.
Co-authored-by: Alex Rickabaugh <alxhub@users.noreply.github.com>
PR Close#50594
Refactor attribute and property binding ingestion and add an attribute extraction phase
Co-authored-by: Alex Rickabaugh <alxhub@users.noreply.github.com>
Co-authored-by: Dylan Hunn <dylhunn@users.noreply.github.com>
Only add the value to the ElementAttributes map for style and attribute kinds
Other kinds should not have their value represented in the consts array
Add missing attribute ingesiton for templates
Unify how template and element bindings are ingested
This resolves the issue of missing listener attributes on templates. In
order to avoid emitting extraneous instructions, listener ops on
templates are stripped in the attribute extraction phase instead.
Handle different binding types separately in ingest
Cleanup code and comments
Disable test that fails on new explicit error.
Previously the test was passing because ingestPropertyBinding treated
attribute bindings as normal bindings which happened to be ok for the
particular test. Now there's an explicit error that attrbiute bindings
aren't yet handled which causes the test to fail
Address feeback
PR Close#50664
Add a flag to disable specific tests when testing the template pipeline
Mark the currently failing tests
Add the template pipeline tests to CI
Update package.json
Co-authored-by: Paul Gschwendtner <paulgschwendtner@gmail.com>
PR Close#50582
According to the HTML specification most attributes are defined as strings, however some can be interpreted as different types like booleans or numbers. [In the HTML standard](https://html.spec.whatwg.org/multipage/common-microsyntaxes.html#boolean-attributes), boolean attributes are considered `true` if they are present on a DOM node and `false` if they are omitted. Common examples of boolean attributes are `disabled` on interactive elements like `<button>` or `checked` on `<input type="checkbox">`. Another example of an attribute that is defined as a string, but interpreted as a different type is the `value` attribute of `<input type="number">` which logs a warning and ignores the value if it can't be parsed as a number.
Historically, authoring Angular inputs that match the native behavior in a type-safe way has been difficult for developers, because Angular interprets all static attributes as strings. While some recent TypeScript versions made this easier by allowing setters and getters to have different types, supporting this pattern still requires a lot of boilerplate and additional properties to be declared. For example, currently developers have to write something like this to have a `disabled` input that behaves like the native one:
```typescript
import {Directive, Input} from '@angular/core';
@Directive({selector: 'mat-checkbox'})
export class MatCheckbox {
@Input()
get disabled() {
return this._disabled;
}
set disabled(value: any) {
this._disabled = typeof value === 'boolean' ? value : (value != null && value !== 'false');
}
private _disabled = false;
}
```
This feature aims to address the issue by introducing a `transform` property on inputs. If an input has a `transform` function, any values set through the template will be passed through the function before being assigned to the directive instance. The example from above can be rewritten to the following:
```typescript
import {Directive, Input, booleanAttribute} from '@angular/core';
@Directive({selector: 'mat-checkbox'})
export class MatCheckbox {
@Input({transform: booleanAttribute}) disabled: boolean = false;
}
```
These changes also add the `booleanAttribute` and `numberAttribute` utilities to `@angular/core` since they're common enough to be useful for most projects.
Fixes#8968.
Fixes#14761.
PR Close#50420
Adds the necessary compiler changes to support input transform functions. The compiler output has changed in the following ways:
### Directive handler
The directive handler now extracts a reference to the input transform function and it resolves the type of its first parameter. It also asserts that the type can be referenced in the compiled output and that it doesn't clash with any pre-existing `ngAcceptInputType_` members.
### .d.ts
In the generated declaration files the compiler now inserts an `ngAcceptInputType_` member for each input with a `transform` function. The member's type corresponds to the type of the first parameter of the function, e.g.
```typescript
// foo.directive.ts
@Directive()
export class Foo {
@Input({transform: (incomingValue: string) => parseInt(incomingValue)}) value: number;
}
// foo.directive.d.ts
export class Foo {
value: number;
static ngAcceptInputType_value: string;
}
```
### Type check block
If an input has `transform` function, the TCB will use the type of its first parameter for the setter type. This uses the same infrastructure as the `ngAcceptInputType_` members.
### Directive declaration
The generated runtime directive declaration call now includes the `transform` function in the `inputs` map, if the input is being transformed. The function will be picked up by the runtime in the next commit to do the actual transformation.
```typescript
// foo.directive.ts
@Directive()
export class Foo {
@Input({transform: (incomingValue: string) => parseInt(incomingValue)}) value: number;
}
// foo.directive.js
export class Foo {
ɵdir = ɵɵdefineDirective({
inputs: {
value: ['value', 'value', incomingValue => parseInt(incomingValue)]
}
});
}
```
PR Close#50225
Fixes that the host directives feature was incorrectly throwing the conflicting alias error when an aliased binding was being exposed under the same alias.
Fixes#48951.
PR Close#50364
This commit adds the `signals: boolean` property to the internal
directive/component metadata. This does not add it to the public API
yet, as the feature has no internal support other than compiler
detection.
PR Close#49981
NgModules which import standalone components currently list those components
in their injector definitions, because we assume that any standalone
component may export providers from its own imports.
This commit adds an optimization for that emit, which attempts to statically
analyze the NgModule imports and determine which standalone components, if
any are present, do not export providers and thus can be omitted.
This analysis is imperfect, because some imported components may be declared
outside of the current compilation, or transitively import types which are
declared outside the compilation. These types are therefore _assumed_ to
carry providers and so the optimization isn't applied to them.
PR Close#49837
The compiler currently does not check to make sure that directives in
the host bindings are exported. These directives are part of the public
API of the component so they do have to be.
PR Close#49527
Adds support for marking a directive input as required. During template type checking, the compiler will verify that all required inputs have been specified and will raise a diagnostic if one or more are missing. Some specifics:
* Inputs are marked as required by passing an object literal with a `required: true` property to the `Input` decorator or into the `inputs` array.
* Required inputs imply that the directive can't work without them. This is why there's a new check that enforces that all required inputs of a host directive are exposed on the host.
* Required input diagnostics are reported through the `OutOfBandDiagnosticRecorder`, rather than generating a new structure in the TCB, because it allows us to provide a better error message.
* Currently required inputs are only supported during AOT compilation, because knowing which bindings are present during JIT can be tricky and may lead to increased bundle sizes.
Fixes#37706.
PR Close#49468
This reverts commit 13dd614cd1.
This breaks a g3 Typescript compilation tests where diagnostics are
expected for a missing input in the component.
PR Close#49467
Adds support for marking a directive input as required. During template type checking, the compiler will verify that all required inputs have been specified and will raise a diagnostic if one or more are missing. Some specifics:
* Inputs are marked as required by passing an object literal with a `required: true` property to the `Input` decorator or into the `inputs` array.
* Required inputs imply that the directive can't work without them. This is why there's a new check that enforces that all required inputs of a host directive are exposed on the host.
* Required input diagnostics are reported through the `OutOfBandDiagnosticRecorder`, rather than generating a new structure in the TCB, because it allows us to provide a better error message.
* Currently required inputs are only supported during AOT compilation, because knowing which bindings are present during JIT can be tricky and may lead to increased bundle sizes.
Fixes#37706.
PR Close#49453
This reverts commit 1a6ca68154.
This breaks tests in google3 which might be depending on private APIs. We
need to update these tests before we can land this PR.
PR Close#49449
Adds support for marking a directive input as required. During template type checking, the compiler will verify that all required inputs have been specified and will raise a diagnostic if one or more are missing. Some specifics:
* Inputs are marked as required by passing an object literal with a `required: true` property to the `Input` decorator or into the `inputs` array.
* Required inputs imply that the directive can't work without them. This is why there's a new check that enforces that all required inputs of a host directive are exposed on the host.
* Required input diagnostics are reported through the `OutOfBandDiagnosticRecorder`, rather than generating a new structure in the TCB, because it allows us to provide a better error message.
* Currently required inputs are only supported during AOT compilation, because knowing which bindings are present during JIT can be tricky and may lead to increased bundle sizes.
Fixes#37706.
PR Close#49304
The decorator downlevel transform is never used for actual class
decorators because Angular class decorators rely on immediate execution
for JIT. Initially we also supported downleveling of class decorators
for View Engine library output, but libraries are shipped using partial
compilation output and are not using this transform anymore.
The transform is exclusively used for JIT processing, commonly for
test files to help ease temporal dead-zone/forward-ref issues. We can
remove the class decorator downlevel logic to remove technical debt.
PR Close#49351
Consider the following scenario:
1. A TS file with a component and templateUrl exists
2. The template file does not exist.
3. First build: ngtsc will properly report the error, via a FatalDiagnosticError
4. The template file is now created
5. Second build: ngtsc still reports the same errror.
ngtsc persists the analysis data of the component and never invalidates
it when the template/style file becomes available later.
This breaks incremental builds and potentially common workflows
where resource files are added later after the TS file is created. This
did surface as an issue in the Angular CLI yet because Webpack requires
users to re-start the process when a new file is added. With ESBuild
this will change and this also breaks incremental builds with
Bazel/Blaze workers.
To fix this, we have a few options:
* Invalidate the analysis when e.g. the template file is missing. Never
caching it means that it will be re-analyzed on every build iteration.
* Add the resource dependency to ngtsc's incremental file graph. ngtsc
will then know via `host.getModifiedResources` when the file becomes
available- and fresh analysis of component would occur.
The first approach is straightforward to implement and was chosen here.
The second approach would allow ngtsc to re-use more of the analysis
when we know that e.g. the template file still not there, but it
increases complexity unnecessarily because there is no **single**
obvious resource path for e.g. a `templateUrl`. The URL is attempted
to be resolved using multiple strategies, such as TS program root dirs,
or there is support for a custom resolution through
`host.resourceNameToFileName`.
It would be possible to determine some candidate paths and add them to
the dependency tracker, but it seems incomplete given possible external
resolvers like `resourceNameToFileName` and also would likely not have
a sufficient-enough impact given that a broken component decorator is
not expected to remain for too long between N incremental build
iterations.
PR Close#49184
The options to generate NgFactory and NgSummary files were added to Ivy for backwards compatibility with ViewEngine. Since ViewEngine was deprecated and removed, the NgFactory and NgSummary files are no longer used as well.
This commit drops obsolete options to generate NgFactory and NgSummary files. Also, the logic that generates those files is also removed.
PR Close#48268
Fixes that the expression converter was producing code that throws a runtime error if a non-null assertion is used as a part of a safe read, write or call.
Fixes#48742.
PR Close#48801
Reworks some of the existing compiler APIs to make them easier to use in a schematic and exposes a few new ones to surface information we already had. High-level list of changes:
* `getPotentialImportsFor` now requires a class reference, instead of a `PotentialDirective | PotentialPipe`.
* New `getNgModuleMetadata` method has been added to the type checker.
* New `getPipeMetadata` method has been added to the type checker.
* New `getUsedDirectives` method has been added to the type checker.
* New `getUsedPipes` method has been added to the type checker.
* The `decorator` property was exposed on the `TypeCheckableDirectiveMeta`. The property was already present at runtime, but it wasn't specified on the interface.
PR Close#48730
In #47167 an `updateClassDeclaration` call was swapped out with a `createClassDeclaration` which caused a regression where interface references were being retained when using a custom decorator in a project that has `emitDecoratorMetadata` enabled.
These changes switch back to use `updateClassDeclaration`.
Fixes#48448.
PR Close#48638
Allows for self-closing tags to be used for non-native tag names, e.g. `<foo [input]="bar"></foo>` can now be written as `<foo [input]="bar"/>`. Native tag names still have to have closing tags.
Fixes#39525.
PR Close#48535
This is basically a pre-step for combining devmode and prodmode into a
single compilation. We are already achieving this now, and can claim
with confidence that we reduced possible actions by half. This is
especially important now that prodmode is used more often, but rules
potentially still using the devmode ESM sources. We can avoid double
compilations (which existed before the whole ESM migration too!).
We will measure this more when we have more concrete documentation
of the changes & a better planning document.
Changes:
* ts_library will no longer generate devmode `d.ts`. Definitions are
generated as part of prodmode. That way only prodmode can be exposed
via providers.
* applied the same to `ng_module`.
* updates migrations to bundle because *everything* using `ts_library`
is now ESM. This is actually also useful in the future if
schematics rely on e.g. the compiler.
* updates schematics for localize to also bundle. similar reason as
above.
PR Close#48521
Since the Bazel setup in this repo will now always use ESM,
the tooling scripts/binaries in AIO need to be switched to ESM
too. Most of the scripts are already ESM, but a few had to be converted.
Note that the Dgeni generation does not use ESM because it's unaffected
and the Dgeni CLI is used. In the future we could also update the Dgeni
setup to ESM but there is no need currently.
PR Close#48521
Since we generate a `.mjs` file as entry-point for jasmine tests,
a couple of issues prevented the transitive dependencies from
bootstrap targets to be brought in (causing resolution errors):
1. The `_files` (previously `_esm2015`) targets are no longer needed,
and they also miss all the information on runfiles.
2. The aspect for computing linker mappings does not respect the
`bootstrap` attribute from the `spec_entrypoint` so we manually
add the extract ESM output targets (this rule works with the aspect
and forwards linker mappings).
PR Close#48521
For every `ts_library` target we expose a shorthand that grants
access to the JS files because `DefaultInfo` of a ts library
only exposes the `.d.ts` files.
We rename this away from `es2015` since in practice it's a much
higher target these days. Additionally we no longer use the devmode
output but rather use the prodmode output which has the explicit
`.mjs` output- compatible with ESM.
PR Close#48521
For standalone components it may be beneficial to group multiple declarations
into a single array, that can then be imported all at once in `Component.imports`.
If this array is declared within a library, however, would the AOT compiler
need to extract the contents of the array from the declaration file. This
requires that the array is constructed using an `as const` cast, which results
in a readonly tuple declaration in the generated .d.ts file of the library:
```ts
export declare const DECLARATIONS: readonly [typeof StandaloneDir];
```
The partial evaluator logic did not support this syntax, so this pattern was
not functional when a library is involved. This commit adds the necessary
logic in the static interpreter to evaluate this type at compile time.
Closes#48089
PR Close#48091
Because the language service uses the compiler, we try to produce as
much useful information as possible rather than throwing hard errors.
Hard errors cause the compiler to crash. While this can be acceptable
when compiling a program as part of a regular build, this is undesirable
for the language service.
PR Close#48314
The stricter checks under `strictInjectionParameters` in Angular 15 now enforce that
an inherited constructor must be compatible with DI, based on whether all parameters
are valid injection tokens. There is an issue when the constructor is inherited from
a class in a declaration file though, as information on the usage of `@Inject()` is
not present within a declaration file. This means that this stricter check cannot be
accurately performed, resulting in false positives.
This commit disables the stricter check to behave the same as it did prior to
Angular 15, therefore avoiding the false positive.
Fixes#48152
PR Close#48156
`getPotentialPipes` returns possible pipes which can be used in the provided context, whether already in scope or requiring an import.
This is necessary to implement auto-import support for pipes in the language service.
PR Close#48090
This commit updates the logic related to the attribute and property binding rules for <iframe> elements. There is a set of <iframe> attributes that may affect the behavior of an iframe and this change enforces that these attributes are only applied as static attributes, making sure that they are taken into account while creating an <iframe>.
If Angular detects that some of the security-sensitive attributes are applied as an attribute or property binding, it throws an error message, which contains the name of an attribute that is causing the problem and the name of a Component where an iframe is located.
BREAKING CHANGE:
Existing iframe usages may have security-sensitive attributes applied as an attribute or property binding in a template or via host bindings in a directive. Such usages would require an update to ensure compliance with the new stricter rules around iframe bindings.
PR Close#47964
This reverts commit 2d08965b1a.
The reason for revert is that we've identified some issues with implementation. The issues will get addressed soon and the fix would be re-submitted.
PR Close#47959
This commit updates the logic related to the attribute order on iframes and makes the rules more strict. There is a set of iframe attributes that may affect the behavior of an iframe, this change enforces that these attributes are applied before an `src` or `srcdoc` attributes are applied to an iframe, so that they are taken into account.
If Angular detects that some of the attributes are set after the `src` or `srcdoc`, it throws an error message, which contains the name of ann attribute that is causing the problem and the name of a Component where an iframe is located. In most cases, it should be enough to change the order of attributes in a template to move the `src` or `srcdoc` ones to the very end.
BREAKING CHANGE:
Existing iframe usages may have `src` or `srcdoc` preceding other attributes. Such usages may need to be updated to ensure compliance with the new stricter rules around iframe bindings.
PR Close#47935
Implements more of the runtime validations for host directives as compiler diagnostics so that they can be caught earlier. Also does some minor cleanup.
PR Close#47768
In AOT compilations, the `strictInjectionParameters` compiler option can
be enabled to report errors when an `@Injectable` annotated class has a
constructor with parameters that do not provide an injection token, e.g.
only a primitive type or interface.
Since Ivy it's become required that any class with Angular behavior
(e.g. the `ngOnDestroy` lifecycle hook) is decorated using an Angular
decorator, which meant that `@Injectable()` may need to have been added
to abstract base classes. Doing so would then report an error if
`strictInjectionParameters` is enabled, if the abstract class has an
incompatible constructor for DI purposes. This may be fine though, as
a subclass may call the constructor explicitly without relying on
Angular's DI mechanism.
Therefore, this commit excludes abstract classes from the
`strictInjectionParameters` check. This avoids an error from being
reported at compile time. If the constructor ends up being used by
Angular's DI system at runtime, then the factory function of the
abstract class will throw an error by means of the `ɵɵinvalidFactory`
instruction.
In addition to the runtime error, this commit also analyzes the inheritance
chain of an injectable without a constructor to verify that their inherited
constructor is valid.
BREAKING CHANGE: Invalid constructors for DI may now report compilation errors
When a class inherits its constructor from a base class, the compiler may now
report an error when that constructor cannot be used for DI purposes. This may
either be because the base class is missing an Angular decorator such as
`@Injectable()` or `@Directive()`, or because the constructor contains parameters
which do not have an associated token (such as primitive types like `string`).
These situations used to behave unexpectedly at runtime, where the class may be
constructed without any of its constructor parameters, so this is now reported
as an error during compilation.
Any new errors that may be reported because of this change can be resolved either
by decorating the base class from which the constructor is inherited, or by adding
an explicit constructor to the class for which the error is reported.
Closes#37914
PR Close#44615
Previously when a file was being analyzed to determine if a shim should
be generated, up to two calls to the host `fileExists` function per file
per generator were made. In the default host, each `fileExists` call made
two underlying file system calls. Following these calls, the file was then
read via `getSourceFile`. However, `getSourceFile` will return `undefined`
if the requested file does not exist. As a result, `getSourceFile` can be
used directly to request both potential file names and leverage the return
value to determine if the file does not exist. This avoids the need to call
`fileExists` at all.
PR Close#47682
`getPotentialImportsFor` returns an array of possible imports, including TypeScript module specifier and identifier name, for a requested trait in the context of a given component.
PR Close#47631
`getPotentialTemplateDirectives` returns possible directives which can be used in the provided context, whether already in scope or requiring an import.
This is necessary to implement auto-import support for standalone components in the language service.
PR Close#47561
Using raw objects as a lookup structure will inadvertently find methods defined on
`Object`, where strings are expected. This causes errors downstream when string
operations are applied on functions.
This commit switches over to use `Map`s in the DOM element schema registry to fix
this category of issues.
Fixes#46936
PR Close#47220
This option has no longer any effect as Ivy is the only rendering engine.
BREAKING CHANGE: Angular compiler option `enableIvy` has been removed as Ivy is the only rendering engine.
PR Close#47346
This is the compile-time implementation of the `hostDirectives` feature plus a little bit of runtime code to illustrate how the newly-generated code will plug into the runtime. It works by creating a call to the new `ɵɵHostDirectivesFeature` feature whenever a directive has a `hostDirectives` field. Afterwards `ɵɵHostDirectivesFeature` will patch a new function onto the directive definition that will be invoked during directive matching.
For example, if we take the following definition:
```ts
@Directive({
hostDirectives: [HostA, {directive: HostB, inputs: ['input: alias']}]
})
class MyDir {}
```
Will compile to:
```js
MyDir.ɵdir = ɵɵdefineComponent({
features: [ɵɵHostDirectivesFeature([HostA, {
directive: HostB,
inputs: {
input: "alias"
}
}])]
});
```
The template type checking is implemented during directive matching by adding the host directives applied on the host to the array of matched directives whenever the host is matched in a template.
Relates to #8785.
PR Close#46868
This helper accepts a class for an Angular trait, and returns the NgModule which owns that trait. This will be useful for the language service import project, which needs to edit import arrays on the module.
PR Close#47166
Replaces (almost) all of the usages of the deprecated `getMutableClone` function from TypeScript which has started to log deprecation warnings in version 4.8 and will likely be removed in version 5.0. The one place we have left is in the default import handling of ngtsc which will be more difficult to remove.
PR Close#47167
This helper accepts a class, and returns the primary Angular Decorator associated with that trait (e.g. the Component, Pipe, Directive, or NgModule decorator). This will be useful for the language service import project, which needs to edit import arrays inside the decorator.
PR Close#47180
Adds support for TypeScript 4.8 and resolves some issues that came up as a result of the update.
Most of the issues came from some changes in TypeScript where the `decorators` and `modifiers` properties were removed from most node types, and were combined into a single `modifiers` array. Since we need to continue supporting TS 4.6 and 4.7 until v15, I ended up creating a new `ngtsc/ts_compatibility` directory to make it easier to reuse the new backwards-compatible code.
PR Close#47038
improve the error message for non-standalone components which are not
exported from their module, and that are also imported directly as if
they were standalone
this change simply adds the suggestion to the developer to import the
ngModule instead
resolves#46004
PR Close#46114
The source-map package now requires the
`SourceMapConsumer`/`SourceMapGenerator` classes to be instantiated
asynchronously. This commit updates our tests to account for that.
PR Close#46888
Saves us some bytes by not emitting `providers` in `defineInjector`. While the amount of bytes isn't huge, I think that this change is worthwhile, because `ng generate` currently generates `providers: []` with every `NgModule` which users can forget to remove.
PR Close#46301
When generating .d.ts metadata for NgModules, by default we emit type
references to their declarations, imports, and exports. However, this
information is not necessarily useful to consumers. References to private
directives (those that aren't exported by the NgModule) for example aren't
at all useful as they can only affect other components declared in the
NgModule. References to imports are of limited usefulness - they might be
helpful for an IDE to understand the DI structure of an application, but
aren't at all used by a downstream compiler.
Generating this metadata is not without cost. When an incremental build
system uses changes in inputs to determine when a rebuild is necessary, any
changes in .d.ts files might cause downstream targets to rebuild. If those
.d.ts changes are in the "private" side of the NgModule (imports or non-
exported directives/pipes), then these rebuilds are wholly unnecessary.
This commit introduces the `onlyPublishPublicTypingsForNgModules` flag for
the compiler. When this flag is set, the compiler will filter the emitted
references in NgModule .d.ts output and only reference those directives/
pipes that are exported from the NgModule (its public API surface). Omitting
the flag preserves the existing behavior of emitting all references, both
public and private.
This is especially useful for build systems such as Bazel.
PR Close#45894
This commit updates the error message to use correct info depending on whether a component is standalone or not. Previously we were always referring to @NgModules as a place to fix the issue, but not we also mention @Component when needed (for standalone components).
PR Close#46159
Angular generally supports cycles between components in the same NgModule.
We have a mechanism of moving the component scope declaration into the
NgModule file in this case. This ensures that Angular never itself
introduces an import which creates a cycle.
What happens if the cycle already exists in the user's program, though, is a
bit different. In these cases, the "correct" emit for Angular is to generate
the component scope (whether direct or remote) inside of a closure, to
prevent evaluating the scope's references until module evaluation is
complete and all cyclic imports have been resolved. We don't want to do this
for *all* scopes because the code size cost of emitting a function wrapper
is non-zero.
In this fix, we take the presence of a `forwardRef` in a component's
`imports` or in an NgModule `declarations` or `imports` as a sign that
component scopes emitted into those files need to be protected against
cyclic references. In a future commit, we may introduce a warning or error
if cyclic imports are not protected behind `forwardRef` in these cases, but
this will take some time to implement.
PR Close#46139
This commit improves the reported error when importing e.g. `RouterModule.forRoot()`
from within `Component.imports`. Such import is not supported, as standalone components
can only refer to other standalone entities or NgModules in their `imports` array;
`ModuleWithProviders` are not supported as `Component.imports` is meant to be used
for the compilation scope of the component, _not_ for configuring DI.
Closes#46003
PR Close#46009
update the error message presented during aot compilation when an unrecognized
tag/element is found in a standalone component so that it does not mention
the ngModule anymore
Note: the jit variant is present in PR #45920resolves#45818
PR Close#45919
This commit accounts for the Babel types changes. Some properties
can now also be `undefined` so existing checks/assertions had to
be adjusted to also capture `undefined` (along with `null`).
Additionally, in preparation for a new ECMA proposal, Babel types
seem to have been updated to include private names in object property
keys. This is not necessarily the case for object expressions, but
could be for object patterns (in the future -- when implemented).
More details: https://github.com/babel/babel/pull/14304 and
https://github.com/tc39/proposal-destructuring-private.
PR Close#45967
This commit fixes a small issue in the logic around the calculation of
template scopes for standalone components. These scopes include a
`Reference` for each dependency of a standalone component, which is used to
generate references to that dependency in various contexts.
Previously, the `Reference` used for a dependency was the one generated from
its own metadata. For example, a referenced directive used the `Reference`
that was created when analyzing the directive declaration itself. This still
works, as the compiler is always able to emit a reference to any valid
`Reference`. However, it's not optimal.
The `Reference` which should be used instead is the one generated from
analyzing the standalone component's `imports` array, which has knowledge of
how the dependency is referenced from within the standalone component's file
itself. This allows the compiler to skip creating a new import for the
dependency when emitting the standalone component, and use the existing,
user-authored import instead. This saves on code size and avoids taxing the
bundler with unnecessary imports.
PR Close#46029
The Angular compiler performs cycle detection when generating imports within
component files. This was previously necessary as reifying dependencies
discovered via NgModules into the component output could add imports that
weren't present in the original component and potentially create cycles.
Doing this could cause order-of-execution issues with existing user imports,
so the compiler detects this case and falls back to an alternative way of
specifying component dependencies that doesn't risk creating cycles.
For standalone components, Angular does not need to add new imports to the
component file as the user has already explicitly referenced dependencies
in the `@Component.imports`. As a result, the cycle detection can be
skipped.
Correctly authoring a program with import cycles is always challenging. One
side of a cyclic import will always initially evaluate to `undefined`, and
this can result in errors in the component definition when this happens
within component `imports`.
Our compiler _could_ detect the cycle and choose to wrap the component
dependencies in an automatic closure instead, avoiding any issues with
`undefined` during an eager evaluation. However, this commit makes an active
choice not to do that as it only serves to mask the problems with cyclic
imports. Future refactorings may cause the "other half" of the cycle to
break. Users should instead be aware of the potential problems with cycles
and explicitly defer evaluations with `forwardRef` where needed. This
ensures that future implementations of Angular compilation which may not be
able to automatically detect import cycles and correct accordingly can still
compile such components.
PR Close#46029
The NodeJS Bazel linker does not work well on Windows because there
is no sandboxing and linker processes from different tests will attempt
to modify the same `node_modules`, causing concurrency race conditions
and resulting in flakiness.
PR Close#45872
This commit improves the error message for using `imports` on a component
that isn't set to `standalone: true`. Two concrete improvements are made:
* A related information message is added to the diagnostic which suggests
the fix of adding `standalone: true`.
* The component is marked as poisoned, preventing other errors which might
be caused by an incorrectly configured template scope from being generated
and thus masking the original problem.
Fixes#45850
PR Close#45851
In AOT compilations, the `strictInjectionParameters` compiler option can
be enabled to report errors when an `@Injectable` annotated class has a
constructor with parameters that do not provide an injection token, e.g.
only a primitive type or interface.
Since Ivy it's become required that any class with Angular behavior
(e.g. the `ngOnDestroy` lifecycle hook) is decorated using an Angular
decorator, which meant that `@Injectable()` may need to have been added
to abstract base classes. Doing so would then report an error if
`strictInjectionParameters` is enabled, if the abstract class has an
incompatible constructor for DI purposes. This may be fine though, as
a subclass may call the constructor explicitly without relying on
Angular's DI mechanism.
Therefore, this commit excludes abstract classes from the
`strictInjectionParameters` check. This avoids an error from being
reported at compile time. If the constructor ends up being used by
Angular's DI system at runtime, then the factory function of the
abstract class will throw an error by means of the `ɵɵinvalidFactory`
instruction.
In addition to the runtime error, this commit also analyzes the inheritance
chain of an injectable without a constructor to verify that their inherited
constructor is valid.
Closes#37914
PR Close#44615
Excludes styles that resolve to empty strings from the emitted metadata so that they don't result in empty `<style>` tags at runtime.
Fixes#31191.
PR Close#45459
This commit updates the logic to detect a situation when a standalone component is used in the NgModule-based bootstrap (`@NgModule.bootstrap`). Both AOT and JIT compilers are updated to handle this situation.
PR Close#45825
In v14, the partial compilation output of components have changed in a way
that prevents older versions of Angular to compile the partial declarations
correctly.
In particular, we used to emit used directives/components in separate arrays called
`components` and `directives`, and used pipes in a property called `pipes`:
```js
TestComponent.ɵcmp = i0.ɵɵngDeclareComponent({
minVersion: "12.0.0",
version: "13.3.0",
type: TestComponent,
selector: "ng-component",
ngImport: i0,
template: ``,
isInline: true,
directives: [{ type: i1.SomeDir, selector: "[some-dir]" }],
components: [{ type: i1.SomeCmp, selector: "some-cmp" }],
pipes: { 'async': i2.AsyncPipe },
});
```
In the above example, the `version` property indicates which exact compiler
version was used to compile the component, but the `minVersion` allows older
versions of the compiler/Angular linker to "link" the partial declaration to
its final AOT compilation output.
In v14, the used directives, components and pipes are now emitted together
into a single array under the `dependencies` property:
```js
TestComponent.ɵcmp = i0.ɵɵngDeclareComponent({
minVersion: "12.0.0",
version: "13.3.0",
type: TestComponent,
selector: "ng-component",
ngImport: i0,
template: ``,
isInline: true,
dependencies: [
{ kind: "directive", type: i1.SomeDir, selector: "[some-dir]" },
{ kind: "component", type: i1.SomeCmp, selector: "some-cmp" },
{ kind: "pipe", type: i2.AsyncPipe },
],
});
```
This change has been made in support of standalone components, but it does mean
that older compiler versions can no longer link these partial declarations
as desirable as none of the components, directives and pipes would be included
in the AOT-compiled code.
By increasing the `minVersion` property, we hint to older compiler versions that
they are not capable of processing the partial declaration. This allows reporting
an error at compile time instead of resulting in runtime failures due to missing
components, directives and pipes.
PR Close#45782
Adds support for TypeScript 4.7. Changes include:
* Bumping the TS version as well as some Bazel dependencies to include https://github.com/bazelbuild/rules_nodejs/pull/3420.
* Adding a backwards-compatibility layer for calls to `updateTypeParameterDeclaration`.
* Making `LView` generic in order to make it easier to type the context based on the usage. Currently the context can be 4 different types which coupled with stricter type checking would required a lot of extra casting all over `core`.
* Fixing a bunch of miscellaneous type errors.
* Removing assertions of `ReferenceEntry.isDefinition` in a few of the language service tests. The field isn't returned by TS anymore and we weren't using it for anything.
* Resolving in error in the language service that was caused by TS attempting to parse HTML files when we try to open them. Previous TS was silently setting them as `ScriptKind.Unknown` and ignoring the errors, but now it throws. I've worked around it by setting them as `ScriptKind.JSX`.
PR Close#45749
Before standalone, everything that could appear in an NgModule's `imports`
was relevant to DI, and needed to be emitted in the `imports` of the
generated `InjectorDef` definition. With the introduction of standalone
types, NgModule `imports` can now contain components, directives, and pipes
which are standalone. Only standalone components need to be included in
the `imports` of the generated injector definition - directives and pipes
have no effect on DI. Having them present doesn't cause any errors in the
runtime (they're filtered out by the injector itself) but it does prevent
tree-shaking.
With this commit, the generation of `InjectorDef` now filters the `imports`
to exclude directives and pipes as much as possible. It's not _always_
possible because an expression in `imports` may pull in both a directive and
a `ModuleWithProviders` reference, and we have no way of referencing just
the MWP part of that expression. Therefore this is an optimization, not a
rule of `InjectorDef` compilation.
PR Close#45701
Previously, the NgModule handler would resolve the `imports` field as one
unit, producing an array of `Reference`s. With this refactoring, if
`imports` is a literal array, each individual element will be resolved
independently. This will allow filtering in the future at the element level,
since there will be a separate `ts.Expression` for each individual element.
PR Close#45701
This commit bundles tests for standalone components that are possible after
previous implementation commits. Most new tests are compliance tests, but
a test is also included to validate that the template type-checking system
can work with standalone components as well.
PR Close#45672
This commit adds a type field to .d.ts metadata for directives, components,
and pipes which carries a boolean literal indicating whether the given type
is standalone or not. For backwards compatibility, this flag defaults to
`false`.
Tests are added to validate that standalone types coming from .d.ts files
can be correctly imported into new standalone components.
PR Close#45672
This commit propagates the `isStandalone` flag for a component, directive,
or pipe during partial compilation of a standalone declaration. This flag
allows the linker to properly process a standalone declaration that it
encounters.
PR Close#45672
Standalone component scopes were first implemented in the
`ComponentDecoratorHandler` itself, due to an assumption that "standalone"
allowed for a localized analysis of the component's dependencies. However,
this is not strictly true. Other compiler machinery also needs to understand
component scopes, including standalone component scopes. A good example is
the template type-checking engine, which uses a `ComponentScopeReader` to
build full metadata objects (that is, metadata that considers the entire
inheritance chain) for type-checking purposes. Therefore, the
`ComponentScopeReader` should be able to give the scope for a standalone
component.
To achieve this, a new `StandaloneComponentScopeReader` is implemented, and
the return type of `ComponentScopeReader.getScopeForComponent` is expanded
to express standalone scopes. This cleanly integrates the "standalone"
concept into the existing machinery.
PR Close#45672
This commit expands on the unified dependency tracking in the previous
commit and adds tracking of NgModule dependencies. These are not used for
standard components, but are emitted for standalone components to allow the
runtime to roll up providers from those NgModules into standalone injectors.
PR Close#45672
Previously, the compiler tracked directives and pipes in template scopes
separately. This commit refactors the scope system to unify them into a
single data structure, disambiguated by a `kind` field.
PR Close#45672
Previously, the compiler would represent template dependencies of a
component in its component definition through separate fields (`directives`,
`pipes`).
This commit refactors the compiler/runtime interface to use a single field
(`dependencies`). The runtime component definition object still has separate
`directiveDefs` and `pipeDefs`, which are calculated from the `dependencies`
when the definition is evaluated.
This change is also reflected in partially compiled declarations. To ensure
compatibility with partially compiled code already on NPM, the linker
will still honor the old form of declaration (with separate fields).
PR Close#45672
This links back each placeholder in a message to the original Angular template span which defines its expression. This is useful for understanding where each placeholder comes from in the context of the full message.
PR Close#45606
This commit carries the `standalone` flag forward from a directive/pipe
into its generated directive/pipe definition, allowing the runtime to
recognize standalone entities.
PR Close#44973
This commit implements the next step of Angular's "standalone" functionality,
by allowing directives/components/pipes declared as `standalone` to be imported
into NgModules. Errors are raised when such a type is not standalone but is
included in an NgModule's imports.
PR Close#44973
This commit improves the error messages generated by the compiler when NgModule
scope analysis finds structural issues within a compilation. In particular,
errors are now shown on a node within the metadata of the NgModule which
produced the error, as opposed to the node of the erroneous declaration/import/
export. For example, if an NgModule declares `declarations: [FooCmp]` and
`FooCmp` is not annotated as a directive, component, or pipe, the error is now
shown on the reference to `FooCmp` in the `declarations` array expression.
Previously, the error would have been shown on `FooCmp` itself, with a mention
in the error text of the NgModule name.
Additional error context in some cases has been moved to related information
attached to the diagnostic, which further improves the legibility of such
errors. Error text has also been adjusted to be more succinct, since more info
about the error is now delivered through context.
PR Close#44973
Recent changes in `rules_nodejs` caused the test case copy file actions
to be transitioned into the `exec` configuration, resulting in much larger
file paths. These paths break on Windows with the shell argument limit, and
with the path limit, causing errors like:
```
ERROR: C:/users/circleci/ng/packages/compiler-cli/test/compliance/test_cases/BUILD.bazel:9:12: Copying file packages/compiler-cli/test/compliance/test_cases/r3_compiler_compliance/components_and_directives/value_composition/structural_directives_if_directive_def.js failed: (Exit 1): cmd.exe failed: error executing command
cd /d C:/users/circleci/_bazel_circleci/u4uoan2j/execroot/angular
SET PATH=C:\Program Files\Git\usr\bin;C:\Program Files\Git\bin;C:\Windows;C:\Windows\System32;C:\Windows\System32\WindowsPowerShell\v1.0
SET RUNFILES_MANIFEST_ONLY=1
cmd.exe /C bazel-out\x64_windows-opt-exec-2B5CBBC6\bin\packages\compiler-cli\test\compliance\test_cases\test_cases--1973427149-cmd.bat
The system cannot find the path specified
```
https://app.circleci.com/pipelines/github/angular/angular/44038/workflows/4b530cb2-f232-4e1d-b35a-e6e085151d08/jobs/1140017
PR Close#45431
.substr() is deprecated so we replace it with functions which work similarily but aren't deprecated
Signed-off-by: Tobias Speicher <rootcommander@gmail.com>
PR Close#45397
When we have an event listener inside an embedded view, we generate a `restoreView` call which saves the view inside of the LFrame. The problem is that we don't clear it until it gets overwritten which can lead to memory leaks.
These changes rework the generated code in order to generate a `resetView` call which will clear the view from the LFrame.
Fixes#42848.
PR Close#43075
In early versions of Angular, it was sometimes necessary to provide a
`moduleId` to `@Component` metadata, and the common pattern for doing this
was to set `moduleId: module.id`. This relied on the bundler to fill in a
value for `module.id`.
However, due to the superficial similarity between `Component.moduleId` and
`NgModule.id`, many users ended up setting `id: module.id` in their
NgModules. This is an anti-pattern that has a few negative effects,
including preventing the NgModule from tree-shaking properly.
This commit changes the compiler to ignore `id: module.id` in NgModules, and
instead provide a warning which suggests removing the line entirely.
PR Close#45024
Angular contains an NgModule registry, which allows a user to declare
NgModules with string ids and retrieve them via those ids, using the
`getNgModuleById` API.
Previously, we attempted to structure this registration in a clever fashion
to allow for tree-shaking of registered NgModules (that is, those with ids).
This sort of worked due to the accidental alignment of behaviors from the
different tree-shakers involved. However, this trick relies on the
generation of `.ngfactory` files and how they're specifically processed in
various bundling scenarios. We intend to remove `.ngfactory` files, hence
we can no longer rely on them in this way.
The correct solution here is to recognize that `@NgModule({id})` is
inherently declaring a global side-effect, and such classes should not
really be eligible for tree-shaking in the first place. This commit removes
all the old registration machinery, and standardizes on generating a side-
effectful call to `registerNgModuleType` for NgModules that have ids.
There is some risk here that NgModules with unnecessary `id`s may not
tree-shake as a result of this change, whereas they would have in previous
circumstances. The fix here should be to remove the `id` if it's not needed.
Specifying an `id` is a request that the NgModule be retained regardless of
any other references, in case it is later looked up by string id.
PR Close#45024
Jasmine logs a warning when there's a `describe` with no tests. These changes fix one such case in the compiler that happens when the tests are run against Windows.
PR Close#45285
The `@angular/localize` package depends on a version of Babel that is two years
old, so this commit updates to the latest version.
Some changes were made to the linker and compliance tests to account for slight
changes in source maps, along with a few code updates because of changes to
the typings of Babel.
PR Close#44931
In templates with several levels of nested nodes, it's common for several `elementStart`/`elementEnd` instructions to show up in a row which can be optimized away.
These changes add chaining support for `elementStart`, `elementEnd`, `elementContainerStart` and `elementContainerEnd` to shave off some bytes when possible.
PR Close#44994
This commit implements the first phase of standalone components in the Angular
compiler. This mainly includes the scoping rules for standalone components
(`@Component({imports})`).
Significant functionality from the design is _not_ implemented by this PR,
including:
* imports of standalone components into NgModules.
* the provider aspect of standalone components
Future commits will address these issues, as we proceed with the design of
this feature.
PR Close#44812
Refs http://b/214103351.
This happens if a user writes `<span i18n>Message</span>`. This is accepted as an internationalized message, but without a description. JSCompiler will throw an error in this situation because descriptions are generally required. Now, the Angular compiler will generate a suppression annotation so JSCompiler allows the syntax. This will ease an internal migration to JSCompiler-based i18n.
PR Close#44787
Attempting to run as is fails because we have `"type": "module"`. `shelljs` is a CommonJS module however, so we need to do a default import and destructure.
```
$ node packages/compiler-cli/test/compliance/update_all_goldens.js
const {exec} = require('shelljs');
^
ReferenceError: require is not defined in ES module scope, you can use import instead
This file is being treated as an ES module because it has a '.js' file extension and '/home/douglasparker/Source/ng/packages/compiler-cli/package.json' contains "type": "module". To treat it as a CommonJS script, rename it to use the '.cjs' file extension.
at file:///home/douglasparker/Source/ng/packages/compiler-cli/test/compliance/update_all_goldens.js:11:16
at ModuleJob.run (node:internal/modules/esm/module_job:183:25)
at async Loader.import (node:internal/modules/esm/loader:178:24)
at async Object.loadESM (node:internal/process/esm_loader:68:5)
at async handleMainPromise (node:internal/modules/run_main:63:12)
```
PR Close#44916
An `ng-template` with an inline template (i.e. has a structural
directive) would previously not get an `undefined` `tagName` because the
logic assumed the element would be `t.Element` or `t.Content` and read
the tag name from the `name` property. For a `t.Template`, this exists
instead on the `t.tagName`. The final result would be an `tagName` of `undefined`
for the parent `t.Template`, causing failures in the indexer downstream.
This `undefined` value is actually expected in the renderer code, even
though the type does not specify this possibility. This change updates
the type of `tagName` to be `string|null` and explicitly handles the
case where there is a structural directive on an `ng-template`. You can
see how the two are differentiated in the compliance code that was
modified in this commit.
PR Close#44788
Previously, if a bad extended diagnostic category was given, it would fail with the expected error as well as an unexpected assertion error:
```
$ ng build -c development
✔ Browser application bundle generation complete.
./src/main.ts - Error: Module build failed (from ./node_modules/@ngtools/webpack/src/ivy/index.js):
Error: Unexpected call to 'assertNever()' with value:
test
at /home/douglasparker/Source/ng-new/node_modules/@ngtools/webpack/src/ivy/loader.js:77:18
at processTicksAndRejections (internal/process/task_queues.js:95:5)
./src/polyfills.ts - Error: Module build failed (from ./node_modules/@ngtools/webpack/src/ivy/index.js):
Error: Unexpected call to 'assertNever()' with value:
test
at /home/douglasparker/Source/ng-new/node_modules/@ngtools/webpack/src/ivy/loader.js:77:18
at processTicksAndRejections (internal/process/task_queues.js:95:5)
Error: error NG4004: Angular compiler option "extendedDiagnostics.checks['invalidBananaInBox']" has an unknown diagnostic category: "test".
Allowed diagnostic categories are:
warning
error
suppress
```
The assertion comes from `ExtendedTemplateCheckerImpl`, which expects a well-formed configuration, yet the compiler would construct it even when errors were found. This commit skips constructing and running extended diagnostics if the configuration had errors, which should avoid triggering these assertion errors.
I'm unfortunately not able to actually test this change. The test passes even before the fix because the `ngc` binary and end-to-end tests [don't request diagnostics unless the configuration is considered valid](ed21f5c753/packages/compiler-cli/src/perform_compile.ts (L292-L293)). See [Slack](https://angular-team.slack.com/archives/C4WHZQMRA/p1642641305003800) for more details.
PR Close#44778
Refs #42966.
Extended diagnostics provide additional analysis about Angular templates by emitting warnings for specific patterns known to be error prone or cause developer confusion. Currently, there are two such diagnostics which are enabled by default:
* `invalidBananaInBox` emits a warning if a user writes a two-way binding backwards like `([foo])="bar"`, when they actually wanted `[(foo)]="bar"`.
* `nullishCoalescingNotNullable` emits a warning if a binding attempts to perform nullish coalescing (`??`) on a type which does not include `null` or `undefined`, such as `{{ foo ?? 'bar' }}` where `foo` is defined as `string` instead of `string | null`.
These diagnostics are enabled as warnings by default, but this can be configured in the `tsconfig.json` like so:
```jsonc
{
"angularCompilerOptions": {
"extendedDiagnostics": {
// The categories to use for specific diagnostics.
"checks": {
// Maps check name to its category.
"invalidBananaInBox": "suppress"
},
// The category to use for any diagnostics not listed in `checks` above.
"defaultCategory": "error"
}
}
}
```
Allowed categories for a diagnostic are `warning` (default), `error`, or `suppress`. `warning` emits the diagnostic but allows the compilation to succeed, `error` *will* fail the compilation, while `suppress` will ignore the diagnostic altogether.
The initial release has two diagnostics, and we are hoping to expand this longer term to add more diagnostics and provide additional insight into Angular templates to detect and surface developer mistakes *before* hours of debugging are wasted.
PR Close#44712
Refs #42966.
This validates the `tsconfig.json` options for extended template diagnostics. It verifies:
* `strictTemplates` must be enabled if `extendedDiagnostics` have any explicit configuration.
* `extendedDiagnostics.defaultCategory` must be a valid `DiagnosticCategoryLabel`.
* `extendedDiagnostics.checks` keys must all be real diagnostics.
* `extendedDiagnostics.checks` values must all be valid `DiagnosticCategoryLabel`s.
These include new error codes, each of which prints out the exact property that was the issue and what the available options are to fix them.
It does disallow the config:
```json
{
"angularCompilerOptions": {
"strictTemplates": false,
"extendedDiagnostics": {
"defaultCategory": "suppress"
}
}
}
```
Such a configuration is technically valid and could be executed, but will be rejected by this verification logic. There isn't much reason to ever do this, since users could just drop `extendedDiagnostics` altogether and get the intended effect. This is unlikely to be a significant issue for users, so it is considered invalid for now to keep the implementation simple.
PR Close#44391
Refs #42966.
The `defaultCategory` option is used for any extended template diagnostics which do not have any specific category specified for them. It defaults to `warning`, since that is the most common behavior expected for users. This provides an easy way for users to promote all diagnostics to errors or suppress all diagnostics.
PR Close#44391
Refs #42966.
This updates `TemplateContext` to include a new `makeTemplateDiagnostic()` function which automatically uses the correct diagnostic category for that check. This makes sure that each diagnostic is emitted with the correct category. It also implicitly passes some known values like `component` and `code` to make the extended template diagnostics a little simpler. Diagnostics which are suppressed are never instantiated at all, which acts as a slight performance optimization since any emitted diagnostics would be ignored anyways.
Unfortunately, diagnostics still have access to `ctx.templateTypeChecker.makeTemplateDiagnostic()` to manually create diagnostics with a different category. Both banana in box and nullish coalescing checks include tests to make sure they respect a manually configured category. This convention should hopefully give a reasonable certainty that new diagnostics will use the correct reporting function, even if that is not strictly enforced.
PR Close#44391
Refs #42966.
This moves extended template check factory invocations into the checker itself, where it can provide a more consistent API contract. Factories are called with compiler options and may return a `TemplateCheck` or `undefined` if the current options do not support that check. This allows `nullishCoalescingNotNullable` to disable itself when `strictNullChecks` is disabled without throwing errors. This gives extended template diagnostics a stronger abstraction model to define their behavior with.
PR Close#44391
In certain scenarios, the compiler may have crashed with an
`Unable to write a reference` error which would be particularly hard
to diagnose. One of the primary reasons for this failure is when the
`rootDir` option is configured---typically the case for libraries---
and a source file is imported using a relative import from an external
entry-point. This would normally report TS6059 for the invalid relative
import, but the crash prevents this error from being surfaced.
This commit refactors the reference emit logic to result in an explicit
`Failure` state with a reason as to why the failure occurred. This state
is then used to report a `FatalDiagnosticException`, preventing a hard
crash.
Closes#44414
PR Close#44587
To make our test output i.e. devmode output more aligned
with what we produce in the NPM packages, or to be more
aligned with what Angular applications will usually consume,
the devmode output is switched from ES5 to ES2015.
Additionally various tsconfigs (outside of Bazel) have been
updated to match with the other parts of the build. The rules
are:
ES2015 for test configurations, ES2020 for actual code that will
end up being shipped (this includes the IDE-only tsconfigs).
PR Close#44505
This commit finishes the removal of View Engine from the codebase, deleting
those pieces of @angular/compiler which were only used for VE.
Co-Authored-By: JoostK <joost.koehoorn@gmail.com>
PR Close#44368
This commit does a first-pass removal of the View Engine infrastructure
in compiler-cli. A more in-depth cleanup is necessary and large parts
of the View Engine compiler infrastructure remain within
`@angular/compiler`, this is just a first cleanup step.
PR Close#44269
As a preparation for the removal of the ViewEngine parts in
`compiler-cli`, this commit moves the version number helper functions
up one level such that the whole `diagnostics` subfolder can be removed.
PR Close#44269
Adds support for TypeScript 4.5. Includes the following changes:
* Bumping the package versions.
* Fixing a few calls to `createExportSpecifier` and `createImportSpecifier` that require an extra parameter.
* Adding some missing methods to the TS compiler hosts.
* Fixing an issue in the TS mocks for the ngcc tests where a regex was too agressive and was trying to match a path like `/node_modules/@typescript/lib-es5`.
* Accounting for type-only import specifiers when reporting DI errors (see #43620).
Fixes#43620.
PR Close#44164
The downlevel decorator transform (commonly used in the CLI and other
tooling of the ecosystem for enabling JIT ES2015+), is currently
incorrectly dealing with nested classes.
The transform will accidentally visit nested classes (in a constructor)
multiple times and generate duplicated instances of the `ctorParameters`
fields. This does not sound like an issue at first, but the duplicated
`ctorParameters` fields will miss significant type information that has
been elided by the first visit, resulting in generated code like the
following:
```js
let MyClass = /* @__PURE__ */ __name(class MyClass extends UpgradeNg1ComponentAdapter {
constructor(scope, injector3, elementRef) {
}
}, "MyClass");
MyClass.ctorParameters = () => [
{ type: void 0, decorators: [{ type: Inject, args: [$SCOPE] }] },
{ type: Injector },
{ type: ElementRef }
];
MyClass.ctorParameters = () => [
{ type: void 0 }, // <---- NOTE!
{ type: Injector },
{ type: ElementRef }
];
```
PR Close#44281
Now that the core package has been cleaned up to no longer contain Ivy
switch code, the transform to switch the `PRE_R3` markers to become
`POST_R3` is deleted as well.
PR Close#43891
This commit removes most tests that were designated as only covering View
Engine code. It also removes tag filters from CI and local commands to run
tests.
In a few cases (such as with the packages/compiler tests), this tag was
improperly applied, and certain test cases have been added back running in
Ivy mode.
This commit also empties `@angular/compiler/testing` as it is no longer
necessary (this is safe since compiler packages are not public API). It can
be deleted in the future.
PR Close#43884
When a partially compiled component or directive is "linked" in JIT mode, the body
of its declaration is evaluated by the JavaScript runtime. If a class is referenced
in a query (e.g. `ViewQuery` or `ContentQuery`) but its definition is later in the
file, then the reference must be wrapped in a `forwardRef()` call.
Previously, query predicates were not wrapped correctly in partial declarations
causing the code to crash at runtime. In AOT mode, this code is never evaluated
but instead transformed as part of the build, so this bug did not become apparent
until Angular Material started running JIT mode tests on its distributable output.
This change fixes this problem by noting when queries are wrapped in `forwardRef()`
calls and ensuring that this gets passed through to partial compilation declarations
and then suitably stripped during linking.
See https://github.com/angular/components/pull/23882 and https://github.com/angular/components/issues/23907
PR Close#44113
Currently the TS version checking function interprets a version like `1.2.3-rc.5` as `1.2.NaN` which would allow it to bypass the version checking altogether.
These changes add a little bit more logic to ensure that such versions are handled correctly. There's also an error if we don't manage to parse the version string.
Also it seemed like we never actually ran the version check unit tests, because they didn't have a test target.
PR Close#44109
When a safe method call such as `person?.getName()` is used, the
compiler would generate invalid code if the argument list also contained
a safe method call. For example, the following code:
```
person?.getName(config?.get('title').enabled)
```
would generate
```
let tmp;
ctx.person == null ? null : ctx.person.getName((tmp = tmp) == null ?
null : tmp.enabled)
```
Notice how the call to `config.get('title')` has completely disappeared,
with `(tmp = tmp)` having taken its place.
The issue occurred due to how the argument list would be converted
from expression AST to output AST twice. First, the outer safe method
call would first convert its arguments list. This resulted in a
temporary being allocated for `config.get('title')`, which was stored in
the internal `_resultMap`. Only after the argument list has been
converted would the outer safe method call realize that it should be
guarded by a safe access of `person`, entering the `convertSafeAccess`
procedure to convert itself. This would convert the argument list once
again, but this time the `_resultMap` would already contain the
temporary `tmp` for `config?.get('title')`. Consequently, the safe
method in the argument list would be emitted as `tmp`.
This commit fixes the issue by ensuring that nodes are only converted
once.
Closes#44069
PR Close#44088
This commit re-enables the `perform_watch` test target and updates the
test to run with the Ivy compiler.
Additionally, this target was switched over to use Angular v12 packages
as input to the test, to allow the ViewEngine tests to continue working
with v13 packages which are Ivy-only. This commit reverts those changes
now that View Engine tests are disabled, as it's desirable to test
against local artifacts that are build within the monorepo instead of
depending on NPM packages.
PR Close#43893
Using the tag "view-engine-only" better describes the expected usage of bazel targets with the test. They can
only be run with view engine.
PR Close#43862
Since building with ViewEngine is not longer desired on CI, removing the ivy vs non-ivy testing yarn scripts
is done, informing developers to instead use `yarn test` as all tests should be run using the Ivy complier.
PR Close#43862
This reverts commit bba0a87055.
The reason for rollback: this change is breaking some targets in Google's codebase when there is no attribute value is displayed (attr.aria-label) when translated.
PR Close#43882
While fully dynamic bound properties (and attributes) cannot be marked for localization, properties that only contain interpolation can.
This commit ensure that attribute bindings that only contain interpolation can also be marked for localization.
Closes#43260
PR Close#43815
When extracting i18n messages from templates, ICU messages are split out from the
message that contains them. This can make it difficult in the translation files to match up
the two messages, especially if the ICU is reused in multiple placeholders.
This commit builds on top of the previous one to expose the message ID of ICU messages
from the ICU placeholders as additional metadata in the `$localize` tagged strings.
Now the metablock following any placeholder can also contain the associated ID
delimited from the placeholder name by `@@`.
Fixes#17506
PR Close#43534
Refs #42966.
Previously, a build when emitted one warning and no errors would fail with a non-zero exit code. This is not what users would expect, but had not been an issue before since the compiler did not actually emit any warnings. With upcoming extended template diagnostics and other warnings, this is now a case that needs to be supported. Warnings are printed to `stderr` as before, but `ngc` now exits with code `0` and the build is considered successful.
Implemented this by adding a new `expectedExitCode` parameter to `driveDiagnostics()` which asserts against the real exit code. Most importantly, it does not **require** the the build to pass since any exit code can be given, so it is up to the test to assert this as well as many messages printed to make sure they are acceptable. This is useful for testing warnings and ensuring the build still passes.
PR Close#43673
Similar to the other private entry-points we have added for localize,
bazel or the migrations, we should expose the tooling code through
a dedicated private export. This will make the compiler-cli exports
more consistent and it will become easier for the CLI to export
necessary code.
PR Close#43431
Removes the remaining usages of dynamic require statements in the
package output. Since we declare all shipped packages as strict ESM,
we cannot use dynamic require statements anymore. This commit switches
these usages to actual `import` statements.
Note: Tsickle continues to remain an optional dependency since bundling
does not work with its UMD package output. Also tsickle is rarely used by
consumers, if at all, so bundling does not really provide any significant
value. To continue keeping tsickle optional (since it's still needed by the
`annotateForClosureCompiler` option which is also respected in ngtsc), we
pass-through a tsickle instance as a parameter to `main`. This allows us to
keep the compile functions synchronous without having to refactor the majority
of the watch compilation code, and majority of tests for ngc, ngtsc.
Consumers (like the `ngc` bin entry-point) can then load tsickle based on their
module format. e.g. tsickle can be imported through `require` to keep everything
sync, but in ESM, the dynamic import can be used beforehand to pass `tsickle` to
the `main` function. We can revisit this in the future but for now this does the
trick without exceeding the scope of this commit..
PR Close#43431
As outlined in the previous commit which enabled the `esModuleInterop`
TypeScript compiler option, we need to update all namespace imports
for `typescript` to default imports. This is needed to allow for
TypeScript to be imported at runtime from an ES module.
Similar changes are needed for modules like `semver` where the types incorrectly
suggest named exports that will not exist at runtime when imported from ESM.
This commit refactors all imports to match with the lint rule we have
configured in the previous commit. See the previous commit for more
details on why certain imports have been changed.
A special case are the imports to `@babel/core` and `@babel/types`. For
these a special interop is needed as both default imports, or named
imports break the other module format. e.g default imports would work
well for ESM, but it breaks for CJS. For CJS, the named imports would
only work, but in ESM, only the default export exist. We work around
this for now until the devmode is using ESM as well (which would be
consistent with prodmode and gives us more valuable test results). More
details on the interop can be found in the `babel_core.ts` files (two
interops are needed for both localize/or the compiler-cli).
PR Close#43431
The view engine language-service tests currently rely on the `npm_package` output
that is built locally. They rely on the package output mostly for
compiling test scenarios (with dependencies on e.g. forms), and for
the testing the metadata extraction (testing proper suggestions for VE).
The reliance on these packages becomes problematic with the new Angular
Package Format v13 where no metadata files are shipped. To continue
being able to test View Engine language-service compatibility, we will
use the v12.x framework packages for some of the test scenarios.
PR Close#43431
The View Engine ngc tests currently rely on the `npm_package` output
that is built locally. This becomes problematic with the new Angular
Package Format v13 where no metadata files are shipped. To continue
being able to test View Engine compilation, we will use the v12.x
framework packages for running the View Engine test.
Note: This means that we no longer test metadata extraction directly
for our framework packages, but given that any change to View Engine
will still land in patch, where the VE packaging still occurs, we should
be covered here.
PR Close#43431
Now that `Route.loadChildren` no longer accepts a string, there is no
need for tooling to find all string-based `loadChildren` to setup lazy
imports for them. As a result, the `listLazyRoutes` operation that
enumerates all string-based `loadChildren` occurrences is no longer
needed and is therefore removed from the compiler.
The `listLazyRoutes` API remains on the `Program` interface to avoid
breaking external tools that may be using this method, but those tools
should ultimately move away from using this API.
PR Close#43591
This commit removes the ability to configure lazy routes using a string
for `loadChildren`, together with the supporting classes to load an
`NgModuleFactory` at runtime.
BREAKING CHANGE:
It is no longer possible to use `Route.loadChildren` using a string
value. The following supporting classes were removed from
`@angular/core`:
- `NgModuleFactoryLoader`
- `SystemJsNgModuleFactoryLoader`
The `@angular/router` package no longer exports these symbols:
- `SpyNgModuleFactoryLoader`
- `DeprecatedLoadChildren`
The signature of the `setupTestingRouter` function from
`@angular/core/testing` has been changed to drop its `NgModuleFactoryLoader`
parameter, as an argument for that parameter can no longer be created.
PR Close#43591
When specifying the `deps` array in the `@Injectable` decorator to
inject dependencies into the injectable's factory function, it should
be possible to use an array literal to configure how the dependency
should be resolved by the DI system.
For example, the following example is allowed:
```ts
@Injectable({
providedIn: 'root',
useFactory: a => new AppService(a),
deps: [[new Optional(), 'a']],
})
export class AppService {
constructor(a) {}
}
```
Here, the `'a'` string token should be injected as optional. However,
the AOT compiler incorrectly used the array literal itself as injection
token, resulting in a failure at runtime. Only if the token were to be
provided using `[new Optional(), new Inject('a')]` would it work
correctly.
This commit fixes the issue by using the last non-decorator in the
array literal as the token value, instead of the array literal itself.
Note that this is a loose interpretation of array literals: if a token
is omitted from the array literal then the array literal itself is used
as token, but any decorator such as `new Optional()` would still have
been applied. When there's multiple tokens in the list then only the
last one will be used as actual token, any prior tokens are silently
ignored. This behavior mirrors the JIT interpretation so is kept as is
for now, but may benefit from some stricter checking and better error
reporting in the future.
Fixes#42987
PR Close#43226
Adds support for TypeScript 4.4. High-level overview of the changes made in this PR:
* Bumps the various packages to `typescript@4.4.2` and `tslib@2.3.0`.
* The `useUnknownInCatchVariables` compiler option has been disabled so that we don't have to cast error objects explicitly everywhere.
* TS now passes in a third argument to the `__spreadArray` call inside child class constructors. I had to update a couple of places in the runtime and ngcc to be able to pick up the calls correctly.
* TS now generates code like `(0, foo)(arg1, arg2)` for imported function calls. I had to update a few of our tests to account for it. See https://github.com/microsoft/TypeScript/pull/44624.
* Our `ngtsc` test setup calls the private `matchFiles` function from TS. I had to update our usage, because a new parameter was added.
* There was one place where we were setting the readonly `hasTrailingComma` property. I updated the usage to pass in the value when constructing the object instead.
* Some browser types were updated which meant that I had to resolve some trivial type errors.
* The downlevel decorators tranform was running into an issue where the Closure synthetic comments were being emitted twice. I've worked around it by recreating the class declaration node instead of cloning it.
PR Close#43281
Currently the compiler has three different classes to represent a "call to something":
1. `MethodCall` - `foo.bar()`
2. `SafeMethodCall` - `foo?.bar()`.
3. `FunctionCall` - Any calls that don't fit into the first two classes. E.g. `foo.bar()()`.
There are a few problems with this approach:
1. It is inconistent with the TypeScript AST which only has one node: `CallExpression`.
2. It means that we have to maintain more code, because the various parts of the compiler need to know about three node types.
3. It doesn't allow us to easily implement some new JS features like safe calls (e.g. `foo.bar?.())`).
These changes rework the compiler so that it produces only one node: `Call`. The new node behaves similarly to the TypeScript `CallExpression` whose `receiver` can be any expression.
There was a similar situation in the output AST where we had an `InvokeMethodExpression` and `InvokeFunctionExpression`. I've combined both of them into `InvokeFunctionExpression`.
PR Close#42882
Previously, the way templates were tokenized meant that we lost information
about the location of interpolations if the template contained encoded HTML
entities. This meant that the mapping back to the source interpolated strings
could be offset incorrectly.
Also, the source-span assigned to an i18n message did not include leading
whitespace. This confused the output source-mappings so that the first text
nodes of the message stopped at the first non-whitespace character.
This commit makes use of the previous refactorings, where more fine grain
information was provided in text tokens, to enable the parser to identify
the location of the interpolations in the original source more accurately.
Fixes#41034
PR Close#43132
The lexer now splits encoded entity tokens out from text and attribute value tokens.
Previously encoded entities would be decoded and the decoded value would be
included as part of the text token of the surrounding text. Now the entities
have their own tokens. There are two scenarios: text and attribute values.
Previously the contents of `<div>Hello & goodbye</div>` would be a single
TEXT token. Now it will be three tokens:
```
TEXT: "Hello "
ENCODED_ENTITY: "&", "&"
TEXT: " goodbye"
```
Previously the attribute value in `<div title="Hello & goodbye">` would be
a single text token. Now it will be three tokens:
```
ATTR_VALUE_TEXT: "Hello "
ENCODED_ENTITY: "&", "&"
ATTR_VALUE_TEXT: " goodbye"
```
- ENCODED_ENTITY tokens have two parts: "decoded" and "encoded".
- ENCODED_ENTITY tokens are always preceded and followed by either TEXT tokens
or ATTR_VALUE_TEXT tokens, depending upon the context, even if they represent
an empty string.
The HTML parser has been modified to recombine these tokens to allow this
refactoring to have limited effect in this commit. Further refactorings
to use these new tokens will follow in subsequent commits.
PR Close#43132
Previously, the decorator transformer was annotating the synthesized properties with TS type annotations. However, because it ran after the JSDoc transformer, the TS types were just dropped from the emitted JS. Attempting to move the decorator transformer before the JSDoc transformer causes tsickle crashes because synthetic AST fragments are not attached to a SourceFile node.
PR Close#43021
Return `TemplateDiagnostic` instead of `ts.Diagnostic` when getting the
extended template diagnostics. This makes the integration with the
language service easier. This also fixes the error code and now uses the
`ngErrorCode` for extended template diagnostics.
Refs #42966
PR Close#43134
This commit adds extended template diagnostics end-to-end tests, to make
sure the diagnostics are generated correctly. Template checks are
already tested with unit tests.
Refs #42966
PR Close#43107
Previously with View Engine output, the `enableResourceInlining` option
could be set to inline external templates and styles (also for the
resulting `.metadata.json` files). We want to do the same for the Ivy
compilation pipeline (regardless of the compilation mode). The full
compilation definitions, and partial declarations currently already
inline resources in a way that no external requests need to be made.
Although there is one exception currently. These are the calls for
setting class metadata (for testbed overrides). This commit updates
the set class metadata calls (for both partial and full compilation)
to always inline resources. This means that libraries do not need
to start shipping external styles/templates just for the
`setClassMetadata` calls.
Note: Only doing this for partial compilation has been considered, but
it seems like it would be simpler implementation-wise to do this for
full compilation as well. Given the external resources are already
inlined (through their `ecmp` definitions), it seems acceptable (or
even more aligned) to do the same for the set class metadata calls.
PR Close#43178
The compliance tests can check source-map segments against expectations
encoded into the expectation files. Previously, the encoding of the expected
segment was only delimited by whitespace, but this made it difficult to identify
segments that started or ended with whitespace.
Now these segment expectations are wrapped in double-quotes which makes
it easier to read and understand the expectation files.
PR Close#43129
Export `getSourceCodeForDiagnostic` from `ngtsc/testing` to make it
available for other packages. This will help confirm that the source
code is correct in other tests.
Refs #42966
PR Close#42984
Previously, the way templates were tokenized meant that we lost information
about the location of interpolations if the template contained encoded HTML
entities. This meant that the mapping back to the source interpolated strings
could be offset incorrectly.
Also, the source-span assigned to an i18n message did not include leading
whitespace. This confused the output source-mappings so that the first text
nodes of the message stopped at the first non-whitespace character.
This commit makes use of the previous refactorings, where more fine grain
information was provided in text tokens, to enable the parser to identify
the location of the interpolations in the original source more accurately.
Fixes#41034
PR Close#42062
The compliance tests can check source-map segments against expectations
encoded into the expectation files. Previously, the encoding of the expected
segment was only delimited by whitespace, but this made it difficult to identify
segments that started or ended with whitespace.
Now these segment expectations are wrapped in double-quotes which makes
it easier to read and understand the expectation files.
PR Close#42062
The static interpreter assumed that a foreign function expression would
have to be imported from the absolute module specifier that was used for
the foreign function itself. This assumption does not hold for the
`forwardRef` foreign function resolver, as that extracts the resolved
expression from the function's argument, which is not behind the
absolute module import of the `forwardRef` function.
The prior behavior has worked for the typical usage of `forwardRef`,
when it is contained within the same source file as where the static
evaluation started. In that case, the resulting reference would
incorrectly have an absolute module guess of `@angular/core`, but the
local identifier emit strategy was capable of emitting the reference
without generating an import using the absolute module guess.
In the scenario where the static interpreter would first have to follow
a reference to a different source that contained the `forwardRef` would
the compilation fail. In that case, there is no local identifier
available such that the absolute module emitter would try to locate the
imported symbol from `@angular/core`. which fails as the symbol is not
exported from there.
This commit fixes the issue by checking whether a foreign expression
occurs in the same source file as the call expression. If it does, then
the absolute module specifier that was used to resolve the call
expression is ignored.
Fixes#42865
PR Close#42887
In #41995 the type of `TrackByFunction` was changed such that the
declaration of a `trackBy` function did not cause the item type to be
widened to the `trackBy`'s item type, which may be a supertype of the
iterated type. This has introduced situations where the template type
checker is now reporting errors for cases where a `trackBy` function is
no longer assignable to `TrackByFunction`.
This commit fixes the error by also including the item type `T` in
addition to the constrained type parameter `U`, allowing TypeScript to
infer an appropriate `T`.
Fixes#42609
PR Close#42692
Currently unless a listener inside of an embedded view tries to reference something from the parent view, or if the reference is a local ref, we don't generate the view restoration instructions and we allow for the value to be picked up from the context object in the function parameters. The problem is that the listener is only run during creation mode and the context object may have been swapped out afterwards.
These changes fix the issue by always generating the view restoration instructions for listeners inside templates.
Fixes#42698.
PR Close#42755
Updates the Bazel NodeJS rules to v4.0.0-beta.0. This is necessary
so that the Angular components repo can update, and it's generally
good to stay as up-to-date as possible with the Bazel rules as it's
easy to fall behind, and updating early allows us to discover issues
affecting our tooling earlier (where they are easier to address due to
e.g. potential breaking change policy).
PR Close#42760
Source files that contain directives or components that need an inline
type constructor or inline template type-check block would always be
considered as affected in incremental rebuilds. The inline operations
cause the source file to be updated in the TypeScript program that is
created for template type-checking, which becomes the reuse program
in a subsequent incremental rebuild.
In an incremental rebuild, the source files from the new user program
are compared to those from the reuse program. The updated source files
are not the same as the original source file from the user program, so
the incremental engine would mark the file which needed inline
operations as affected. This prevents incremental reuse for these files,
causing sub-optimal rebuild performance.
This commit attaches the original source file for source files that have
been updated with inline operations, such that the incremental engine
is able to compare source files using the original source file.
Fixes#42543
PR Close#42759
As of ES2021, JavaScript allows using underscores as separators inside numbers, in order to make them more readable (e.g. `1_000_000` vs `1000000`). TypeScript has had support for separators for a while so these changes expand the template parser to handle them as well.
PR Close#42672
Adds support for shorthand property declarations inside Angular templates. E.g. doing `{foo, bar}` instead of `{foo: foo, bar: bar}`.
Fixes#10277.
PR Close#42421
If an implcit receiver is accessed in a listener inside of an `ng-template`, we generate some extra code in order to ensure that we're assigning to the correct object. The problem is that the logic wasn't covering keyed writes which caused it to write to the wrong object and throw an assertion error at runtime.
These changes expand the logic to cover keyed writes.
Fixes#41267.
PR Close#42603
xi18n is the operation of extracting i18n messages from templates in the
compilation. Previously, only View Engine was able to perform xi18n. This
commit implements xi18n in the Ivy compiler, and a copy of the View Engine
test for Ivy verifies that the results are identical.
PR Close#42485
This is based on a discussion we had a few weeks ago. Currently if a component uses `ViewEncapsulation.ShadowDom` and its selector doesn't meet the requirements for a custom element tag name, a vague error will be thrown at runtime saying something like "Element does not support attachShadowRoot".
These changes add a new diagnostic to the compiler that validates the component selector and gives a better error message during compilation.
PR Close#42245
Updates the compiler-cli compliance goldens. The golden updates are
required due to a regression in TypeScript 4.3 that causes the emitter
to not incorrectly preserve lines when emitting a node list.
This can be reverted once https://github.com/microsoft/TypeScript/pull/44070
is available.
PR Close#42022
Switches the repository to TypeScript 4.3 and the latest
version of tslib. This involves updating the peer dependency
ranges on `typescript` for the compiler CLI and for the Bazel
package. Tests for new TypeScript features have been added to
ensure compatibility with Angular's ngtsc compiler.
PR Close#42022
Currently we support safe property (`a?.b`) and method (`a?.b()`) accesses, but we don't handle safe keyed reads (`a?.[0]`) which is inconsistent. These changes expand the compiler in order to support safe key read expressions as well.
PR Close#41911
The `ModuleWithProviders` type has required a generic type since Angular 10,
so it is no longer necessary for the compiler to transform usages of the
`ModuleWithProviders` type without the generic type, as that should have
been reported as a compile error. This commit removes the detection logic
from ngtsc.
PR Close#41996
When a `trackBy` function is used that accepts a supertype of the iterated
array's type, the loop variable would undesirably be inferred as the supertype
instead of the array's item type. This commit adds an inferred type parameter
to `TrackByFunction` to allow an extra degree of freedom, enabling the
loop value to be inferred as the most narrow type.
Fixes#40125
PR Close#41995
The ngtsc test targets have fake declarations files for `@angular/core`
and `@angular/common` and the template type checking tests can leverage
the fake common declarations instead of declaring its own types.
PR Close#41995
Type-only imports are known to be elided by TypeScript, so the compiler
can be certain that such imports do not contribute to potential import
cycles. As such, type-only imports are no longer considered during cycle
analysis.
Regular import statements that would eventually be fully elided by
TypeScript during emit if none of the imported symbols are used in a
value position continue to be included in the cycle analysis, as the
cycle analyzer is unaware of these elision opportunities. Only explicit
`import type` statements are excluded.
PR Close#42453
The compiler flag `compileNonExportedClasses` allows the Angular compiler to
process classes which are not exported at the top level of a source file.
This is often used to allow for AOT compilation of test classes inside
`it()` test blocks, for example.
Previously, the compiler would identify exported classes by looking for an
`export` modifier on the class declaration itself. This works for the
trivial case, but fails for indirectly exported classes:
```typescript
// Component is declared unexported.
@Component({...})
class FooCmp {...}
// Indirect export of FooCmp
export {FooCmp};
```
This is not an immediate problem for most application builds, since the
default value for `compileNonExportedClasses` is `true` and therefore such
classes get compiled regardless.
However, in the Angular Language Service now, `compileNonExportedClasses` is
forcibly overridden to `false`. That's because the tsconfig used by the IDE
and Language Service is often far broader than the application build's
configuration, and pulls in test files that can contain unexported classes
not designed with AOT compilation in mind.
Therefore, the Language Service has trouble working with such structures.
In this commit, the `ReflectionHost` gains a new API for detecting whether a
class is exported. The implementation of this method now not only considers
the `export` modifier, but also scans the `ts.SourceFile` for indirect
exports like the example above. This ensures the above case will be
processed directly in the Language Service.
This new operation is cached using an expando symbol on the `ts.SourceFile`,
ensuring good performance even when scanning large source files with lots of
exports (e.g. a FESM file under `ngcc`).
Fixes#42184.
PR Close#42207
When `checkTypeOfPipes` is set to `false`, our TCB currently generates
the a statement like the following when pipes appear in the template:
`(_pipe1 as any).transform(args)`
This did enable us to get _some_ information from the Language Service
about pipes in this case because we still had access to the pipe
instance. However, because it is immediately cast to `any`, we cannot
get type information about the transform access. That means actions like "go to
definition", "find references", "quick info", etc. will return
incomplete information or fail altogether.
Instead, this commit changes the TCB to generate `(_pipe1.transform as any)(args)`.
This gives us the ability to get complete information for the LS
operations listed above.
PR Close#40523
This commit changes the reference emitters in the Ivy compiler to prefer
non-aliased exports if they exist. This avoids selecting "private
exports" that may not be stable, e.g. the reexports that have been added
by the View Engine compiler. Such reexports are not stable and are
therefore not suitable to be emitted into partial compilations, as the
output of partial compilations should only reference stable symbols
from upstream libraries.
An alternative solution has been considered where ViewEngine-generated
exports would gain a certain prefix, such that the Ivy compiler could
just exclude those exports (see #41443). However, that solution would
be insufficient in case a library is built using partial compilation and
while depending itself on a VE-compiled library from earlier versions of
Angular, where the magic prefix would be missing. For such libraries,
ngcc would have generated reexports using the declared name if not already
present so this change does result in choosing the correct export.
Because ngcc always generates reexports using the declared name even if
an aliased export is present, this change causes those ngcc-generated
exports to be chosen in downstream libraries using partial compilation.
This is unfortunate as it means that the declared names become
effectively public even if the library author was intentionally
exporting it using an alias. This commit does not address this problem;
it is expected that this should not result in widespread issues across
the library ecosystem.
Fixes#41277
PR Close#41866
We have a check that determines whether to generate property binding instructions for an `ng-template`. The check looks at whether the tag name is exactly `ng-template`, but the problem is that if the tag is placed in a non-HTML namespace (e.g. `svg`), the tag name will actually be `:namespace:ng-template` and the check will fail.
These changes resolve the issue by looking at the tag name without the namespace.
Fixes#41308.
PR Close#41669
Currently if a component defines a template inline, but not through a
string literal, the partial compilation references the template expression
as is. This is problematic because the component declaration can no longer
be processed by the linker later as there is no static interpretation. e.g.
```js
const myTemplate = `...`;
TestCmp.ɵcmp = i0.ɵɵngDeclareComponent({
version: "0.0.0-PLACEHOLDER",
type: TestCmp,
selector: "test-cmp",
ngImport: i0,
template: myTemplate,
isInline: true
});
```
To fix this, we use the the resolved template in such cases so that
the linker can process the template/component declaration as expected.
PR Close#41583
This is follow-up from #41437 and it reduces the amount of code we generate for safe property accesses (`a?.b`) and nullish coalescing (`a ?? b`) by:
1. Reusing variables in nested nullish coalescing expressions.
2. Not initializing temporary variables to `null`. The way our code is generated means that the value will always be overwritten before we compare against it so the initializer didn't really matter.
Fixes#41491.
PR Close#41563
Previously, it was not possible to block a partial-linker from trying to
process a declaration that was defined in a newer version of Angular than
that of the partial-linker. For example, if a partial-linker was published as
part of version 12.0.0, there was no way for a partially-compiled declaration
compiled via version 13.0.0 to tell the 12.0.0 linker that it would be invalid
to attempt to process it.
This commit adds a new `minVersion` property to partial-declarations, which is
interpreted as the "minimum partial-linker version" that can process this
declaration. When selecting a partial-linker for such a declaration, the known
linker version ranges are checked to find the most recent linker whose version
range has an overlap with the interpreted declaration range.
This approach allows us to set a minimum version for a declaration, which
can inform an old partial-linker that will it not be able to accurately
process the declaration.
Note that any pre-release part to versions are ignored in this selection
process.
The file-linker can be configured, via the `unknownDeclarationVersionHandling`
property of `LinkerOptions`, to handle such a situation in one of three ways:
- `error` - the version mismatch is a fatal error
- `warn` - a warning is sent to the logger but the most recent partial-linker
will attempt to process the declaration anyway.
- `ignore` - the most recent partial-linker will, silently, attempt to process
the declaration.
The default is to throw an error.
Closes#41497
PR Close#41578
There were three options being made available to users of the linker:
- ` enableI18nLegacyMessageIdFormat`
- `i18nNormalizeLineEndingsInICUs`
- ` i18nUseExternalIds`
None of these should actually be configurable at linking time
because partially-linked libraries have tighter restrictions on
what i18n options can be used.
This commit removes those options from the `LinkerOptions` interface.
It was considered to add a check for backwards compatibilty to ensure
that if these options were being passed, and were different to the expected
defaults, we would throw an informative error. But from looking at the
Angular CLI (the only known client of the linker) it has never been setting
these options so they have already always been set to the defaults.
BREAKING CHANGE:
Linked libraries no longer generate legacy i18n message ids. Any downstream
application that provides translations for these messages, will need to
migrate their message ids using the `localize-migrate` command line tool.
Closes#40673
PR Close#41554
When an Ivy NgModule is imported into a View Engine build, it doesn't have
metadata.json files that describe it as an NgModule, so it appears to VE
builds as a plain, undecorated class. The error message shown in this
situation generic and confusing, since it recommends adding an @NgModule
annotation to a class from a library.
This commit adds special detection into the View Engine compiler to give a
more specific error message when an Ivy NgModule is imported.
PR Close#41534
When multiple occurrences of the same package exist within a single
TypeScript compilation unit, TypeScript deduplicates the source files
by introducing redirected source file proxies. Such proxies are
recreated during an incremental compilation even if the original
declaration file did not change, which caused the compiler not to reuse
any work from the prior compilation.
This commit changes the incremental driver to recognize a redirected
source file and treat them as their unredirected source file.
PR Close#41448
The Angular compiler has to actively keep default import statements
alive if they were only used in type-only positions, but have been
emitted as value expressions for DI purposes. A problem occurred in
incremental recompilations, where the relationship between an identifier
usage and its corresponding default import would not be considered. This
could result in the removal of the default import statement and caused
a `ReferenceError` at runtime.
This commit fixes the issue by storing the association from an
identifier to its default import declaration on the source file itself,
instead of within the `DefaultImportTracker` instance. The
`DefaultImportTracker` instance is only valid for a single compilation,
whereas the association from an identifier to a default import
declaration is valid as long as the `ts.SourceFile` is the same
instance.
A subsequent commit refactor the `DefaultImportTracker` to no longer
be responsible for registering the association, as its lifetime is
conceptually too short to do so.
Fixes#41377
PR Close#41557
The `emitDecoratorMetadata` compiler option does not have to be enabled
as Angular decorators are transformed by the AOT compiler. Having the
option enabled in our tests can hide issues around import preservation,
as with `emitDecoratorMetadata` enabled the TypeScript compiler itself
does not elide imports even if they are only used in type-positions.
This is unlike having `emitDecoratorMetadata` disabled, however; in that
case the Angular compiler has to actively trick TypeScript into
retaining default imports when an identifier in a type-only position has
been reified into a value position for DI purposes.
A subsequent commit addresses a bug in default import preservation
that relies on this flag being `false`.
PR Close#41557
This commit refactors the generated code for class metadata in partial
compilation mode. Instead of emitting class metadata into a top-level
`ɵsetClassMetadata` call guarded by `ngDevMode` flags, the class
metadata is now declared using a top-level `ɵɵngDeclareClassMetadata`
call.
PR Close#41200
This commit changes the partial compilation so that it outputs declarations
rather than definitions for injectables.
The JIT compiler and the linker are updated to be able to handle these
new declarations.
PR Close#41316
The other similar interfaces were renamed in https://github.com/angular/angular/pull/41119,
but this one was left since it had existed before Ivy. It looks like the interface was
never actually exposed on npm so it is safe to rename this one too.
PR Close#41316
Now that other values were removed from `R3ResolvedDependencyType`,
its meaning can now be inferred from the other properties in the
`R3DeclareDependencyMetadata` type. This commit removes this enum
and updates the code to work without it.
PR Close#41231
This instruction was created to work around a problem with injecting a
`ChangeDetectorRef` into a pipe. See #31438. This fix required special
metadata for when the thing being injected was a `ChangeDetectorRef`.
Now this is handled by adding a flag `InjectorFlags.ForPipe` to the
`ɵɵdirectiveInject()` call, which avoids the need to special test_cases
`ChangeDetectorRef` in the generated code.
PR Close#41231
This commit changes the partial compilation so that it outputs declaration
calls rather than compiled factory functions.
The JIT compiler and the linker are updated to be able to handle these
new declarations.
PR Close#41231
The `ɵɵInjectorDef` interface is internal and should not be published publicly
as part of libraries. This commit updates the compiler to render an opaque
type, `ɵɵInjectorDeclaration`, for this instead, which appears in the typings
for compiled libraries.
PR Close#41119
Th `ɵɵFactoryDef` type will appear in published libraries, via their typings
files, to describe what type dependencies a DI factory has. The parameters
on this type are used by tooling such as the Language Service to understand
the DI dependencies of the class being created by the factory.
This commit moves the type to the `public_definitions.ts` file alongside
the other types that have a similar role, and it renames it to `ɵɵFactoryDeclaration`
to align it with the other declaration types such as `ɵɵDirectiveDeclaration`
and so on.
PR Close#41119
These types are only used in the generated typings files to provide
information to the Angular compiler in order that it can compile code
in downstream libraries and applications.
This commit aliases these types to `unknown` to avoid exposing the
previous alias types such as `ɵɵDirectiveDef`, which are internal to
the compiler.
PR Close#41119
The partial declaration of a component includes the list of directives
that are used in its template, including some metadata of the directive
which can be used during actual compilation of the component. Used
components are currently part of this list, as components are also
directives. This commit splits the used components into a dedicate
property in the partial declaration, which allows for template
compilation to optimize the generated code for components.
PR Close#41104
This commit changes the partial compilation so that it outputs declaration
calls rather than definition calls for NgModules and Injectors.
The JIT compiler and the linker are updated to be able to handle these
new declarations.
PR Close#41080
BREAKING CHANGE:
Switching default of `emitDistinctChangesOnlyDefaultValue`
which changes the default behavior and may cause some applications which
rely on the incorrect behavior to fail.
`emitDistinctChangesOnly` flag has also been deprecated and will be
removed in a future major release.
The previous implementation would fire changes `QueryList.changes.subscribe`
whenever the `QueryList` was recomputed. This resulted in an artificially
high number of change notifications, as it is possible that recomputing
`QueryList` results in the same list. When the `QueryList` gets recomputed
is an implementation detail, and it should not be the thing that determines
how often change event should fire.
Unfortunately, fixing the behavior outright caused too many existing
applications to fail. For this reason, Angular considers this fix a
breaking fix and has introduced a flag in `@ContentChildren` and
`@ViewChildren`, that controls the behavior.
```
export class QueryCompWithStrictChangeEmitParent {
@ContentChildren('foo', {
// This option is the new default with this change.
emitDistinctChangesOnly: true,
})
foos!: QueryList<any>;
}
```
For backward compatibility before v12
`emitDistinctChangesOnlyDefaultValue` was set to `false. This change
changes the default to `true`.
PR Close#41121
Previously, injector definitions contained a `factory` property that
was used to create a new instance of the associated NgModule class.
Now this factory has been moved to its own `ɵfac` static property on the
NgModule class itself. This is inline with how directives, components and
pipes are created.
There is a small size increase to bundle sizes for each NgModule class,
because the `ɵfac` takes up a bit more space:
Before:
```js
let a = (() => {
class n {}
return n.\u0275mod = c.Cb({type: n}),
n.\u0275inj = c.Bb({factory: function(t) { return new (t || n) }, imports: [[e.a.forChild(s)], e.a]}),
n
})(),
```
After:
```js
let a = (() => {
class n {}
return n.\u0275fac = function(t) { return new (t || n) },
n.\u0275mod = c.Cb({type: n}),
n.\u0275inj = c.Bb({imports: [[r.a.forChild(s)], r.a]}),
n
})(),
```
In other words `n.\u0275fac = ` is longer than `factory: ` (by 5 characters)
and only because the tooling insists on encoding `ɵ` as `\u0275`.
This can be mitigated in a future PR by only generating the `ɵfac` property
if it is actually needed.
PR Close#41022
This change marks all relevant define* callsites as pure, causing the compiler to
emmit either @__PURE__ or @pureOrBreakMyCode annotation based on whether we are
compiling code annotated for closure or terser.
This change is needed in g3 where we don't run build optimizer but we
need the code to be annotated for the closure compiler.
Additionally this change allows for simplification of CLI and build optimizer as they
will no longer need to rewrite the generated code (there are still other places where
a build optimizer rewrite will be necessary so we can't remove it, we can only simplify it).
PR Close#41096
In Angular programs, changing a file may require other files to be
emitted as well due to implicit NgModule dependencies. For example, if
the selector of a directive is changed then all components that have
that directive in their compilation scope need to be recompiled, as the
change of selector may affect the directive matching results.
Until now, the compiler solved this problem using a single dependency
graph. The implicit NgModule dependencies were represented in this
graph, such that a changed file would correctly also cause other files
to be re-emitted. This approach is limited in a few ways:
1. The file dependency graph is used to determine whether it is safe to
reuse the analysis data of an Angular decorated class. This analysis
data is invariant to unrelated changes to the NgModule scope, but
because the single dependency graph also tracked the implicit
NgModule dependencies the compiler had to consider analysis data as
stale far more often than necessary.
2. It is typical for a change to e.g. a directive to not affect its
public API—its selector, inputs, outputs, or exportAs clause—in which
case there is no need to re-emit all declarations in scope, as their
compilation output wouldn't have changed.
This commit implements a mechanism by which the compiler is able to
determine the impact of a change by comparing it to the prior
compilation. To achieve this, a new graph is maintained that tracks all
public API information of all Angular decorated symbols. During an
incremental compilation this information is compared to the information
that was captured in the most recently succeeded compilation. This
determines the exact impact of the changes to the public API, which
is then used to determine which files need to be re-emitted.
Note that the file dependency graph remains, as it is still used to
track the dependencies of analysis data. This graph does no longer track
the implicit NgModule dependencies, which allows for better reuse of
analysis data.
These changes also fix a bug where template type-checking would fail to
incorporate changes made to a transitive base class of a
directive/component. This used to be a problem because transitive base
classes were not recorded as a transitive dependency in the file
dependency graph, such that prior type-check blocks would erroneously
be reused.
This commit also fixes an incorrectness where a change to a declaration
in NgModule `A` would not cause the declarations in NgModules that
import from NgModule `A` to be re-emitted. This was intentionally
incorrect as otherwise the performance of incremental rebuilds would
have been far worse. This is no longer a concern, as the compiler is now
able to only re-emit when actually necessary.
Fixes#34867Fixes#40635Closes#40728
PR Close#40947
For certain generated function calls, the compiler emits a 'PURE' annotation
which informs Terser (the optimizer) about the purity of a specific function
call. This commit expands that system to produce a new Closure-specific
'pureOrBreakMyCode' annotation when targeting the Closure optimizer instead
of Terser.
PR Close#41021
This is a pre-requisite for #40360. Given the following template which has a listener
that references a variable from a parent template (`name`):
```
<ng-template let-name="name">
<button (click)="hello(name)"></button>
</ng-template>
```
We generate code that looks that looks like. Note how we access `name` through `ctx`:
```js
function template(rf, ctx) {
if (rf & 1) {
const r0 = ɵɵgetCurrentView();
ɵɵelementStart(0, "button", 2);
ɵɵlistener("click", function() {
ɵɵrestoreView(r0);
const name_r0 = ctx.name; // Note the `ctx.name` access here.
const ctx_r1 = ɵɵnextContext();
return ctx_r1.log(name_r0);
});
ɵɵelementEnd();
}
}
```
This works fine at the moment, because the template context object can't be changed after creation.
The changes in #40360 allow for the object to be changed, which means that the `ctx` reference
inside the listener will be out of date, because it was bound during creation mode.
This PR aims to address the issue by accessing the context inside listeners through the saved
view reference. With the new code, the generated code from above will look as follows:
```js
function template(rf, ctx) {
if (rf & 1) {
const r0 = ɵɵgetCurrentView();
ɵɵelementStart(0, "button", 2);
ɵɵlistener("click", function() {
const restoredCtx = ɵɵrestoreView(r0);
const name_r0 = restoredCtx.name;
const ctx_r1 = ɵɵnextContext();
return ctx_r1.log(name_r0);
});
ɵɵelementEnd();
}
}
```
PR Close#40833
Our approach for handling cyclic imports results in code that is
not easy to tree-shake, so it is not suitable for publishing in a
library.
When compiling in partial compilation mode, we are targeting
such library publication, so we now create a fatal diagnostic
error instead of trying to handle the cyclic import situation.
Closes#40678
PR Close#40782
This commit implements creating of `ɵɵngDeclarePipe()` calls in partial
compilation, and processing of those calls in the linker and JIT compiler.
See #40677
PR Close#40803
Produces a diagnostic when we cannot resolve a component's external style sheet or external template.
The previous behavior was to throw an exception, which crashed the
Language Service.
fixes angular/vscode-ng-language-service#1079
PR Close#40660
Normally the template parsing operation normalizes all template line endings
to '\n' only. This normalization operation causes source mapping errors when
the original template uses '\r\n' line endings.
The compiler already parses templates again to create a "diagnostic"
template AST with accurate source maps, to avoid other parsing issues that
affect source map accuracy. This commit configures this diagnostic parse to
also preserve line endings.
PR Close#40597
If the template parse option `leadingTriviaChars` is configured to
consider whitespace as trivia, any trailing whitespace of an element
would be considered as leading trivia of the subsequent element, such
that its `start` span would start _after_ the whitespace. This means
that the start span cannot be used to mark the end of the current
element, as its trailing whitespace would then be included in its span.
Instead, the full start of the subsequent element should be used.
To harden the tests that for the Ivy parser, the test utility `parseR3`
has been adjusted to use the same configuration for `leadingTriviaChars`
as would be the case in its production counterpart `parseTemplate`. This
uncovered another bug in offset handling of the interpolation parser,
where the absolute offset was computed from the start source span
(which excludes leading trivia) whereas the interpolation expression
would include the leading trivia. As such, the absolute offset now also
uses the full start span.
Fixes#39148
PR Close#40513
The compliance test runner has various macros that process the
expectation files before actually checking their contents. Among those
macros are i18n helpers, which uses a global message counter to be able
to uniquely identify ICU variables.
Because of the global nature of this message index, it was susceptible
to ordering issues which could result in flaky tests, although it failed
very infrequently.
This commit resets the global message counter before applying the macros.
As a result of this change an expectation file had to be updated; this
is actually a bug fix as said test used to fail if run in isolation (if
`focusTest: true` was set for that particular testcase).
PR Close#40529
Because the query now has `flags` which specify the mode, the static query
instruction can now be remove. It is simply normal query with `static` flag.
PR Close#40091
Previous implementation would fire changes `QueryList.changes.subscribe`
whenever the `QueryList` was recomputed. This resulted in artificially
high number of change notifications, as it is possible that recomputing
`QueryList` results in the same list. When the `QueryList` gets recomputed
is an implementation detail and it should not be the thing which determines
how often change event should fire.
This change introduces a new `emitDistinctChangesOnly` option for
`ContentChildren` and `ViewChildren`.
```
export class QueryCompWithStrictChangeEmitParent {
@ContentChildren('foo', {
// This option will become the default in the future
emitDistinctChangesOnly: true,
})
foos!: QueryList<any>;
}
```
PR Close#40091
The `template` and `isInline` fields were previously stored in a nested
object, which was initially done to accommodate for additional template
information to support accurate source maps for external templates. In
the meantime the source mapping has been accomplished in a different
way, and I feel this flattened structure is simpler and smaller so is
preferable over the nested object. This change also makes the `isInline`
property optional with a default value of `false`.
PR Close#40383
The parser has a list of tag definitions that it uses when parsing the template. Each tag has a
`contentType` which tells the parser what kind of content the tag should contain. The problem is
that the browser has two separate `title` tags (`HTMLTitleElement` and `SVGTitleElement`) and each
of them has to have a different `contentType`, otherwise the parser will throw an error further down
the pipeline.
These changes update the tag definitions so that each tag name can have multiple content types
associated with it and the correct one can be returned based on the element's prefix.
Fixes#31503.
PR Close#40259
The decorator downleveling transform patches `ts.EmitResolver.isReferencedAliasDeclaration`
to prevent elision of value imports that occur only in a type-position, which would
inadvertently install the patch repeatedly for each source file in the program.
This could potentially result in a stack overflow when a very large number of files is
present in the program.
This commit fixes the issue by ensuring that the patch is only applied once.
This is also a slight performance improvement, as `isReferencedAliasDeclaration`
is no longer repeatedly calling into all prior installed patch functions.
Fixes#40276
PR Close#40374
Now that `ReadonlyFileSystem` and `PathManipulation` interfaces are
available, this commit updates the compiler-cli to use these more
focussed interfaces.
PR Close#40281
Currently when analyzing the metadata of a directive, we bundle together the bindings from `host`
and the `HostBinding` and `HostListener` together. This can become a problem later on in the
compilation pipeline, because we try to evaluate the value of the binding, causing something like
`@HostBinding('class.foo') public true = 1;` to be treated the same as
`host: {'[class.foo]': 'true'}`.
While looking into the issue, I noticed another one that is closely related: we weren't treating
quoted property names correctly. E.g. `@HostBinding('class.foo') public "foo-bar" = 1;` was being
interpreted as `classProp('foo', ctx.foo - ctx.bar)` due to the same issue where property names
were being evaluated.
These changes resolve both of the issues by treating all `HostBinding` instance as if they're
reading the property from `this`. E.g. the `@HostBinding('class.foo') public true = 1;` from above
is now being treated as `host: {'[class.foo]': 'this.true'}` which further down the pipeline becomes
`classProp('foo', ctx.true)`. This doesn't have any payload size implications for existing code,
because we've always been prefixing implicit property reads with `ctx.`. If the property doesn't
have an identifier that can be read using dotted access, we convert it to a quoted one (e.g.
`classProp('foo', ctx['is-foo']))`.
Fixes#40220.
Fixes#40230.
Fixes#18698.
PR Close#40233
This commit changes the `PartialComponentLinker` to use the original source
of an external template when compiling, if available, to ensure that the
source-mapping of the final linked code is accurate.
If the linker is given a file-system and logger, then it will attempt
to compute the original source of external templates so that the final
linked code references the correct template source.
PR Close#40237
Now, if a source-mapping compliance test fails, the message displays both
the path to the generated file, and more helpfully the path to the expected
file.
PR Close#40237
Previously the names of the source and expectation files were often reused,
which caused potential confusion.
There is now a single source file for
each test-case, which is important when they are being compiled with different
compiler options, since the GOLDEN_PARTIAL file will only contain one copy
per file name.
The names of the expectation files have now been changed so that is clearer
which test-case they are related to.
PR Close#40237
The trustConstantHtml and trustConstantResourceUrl functions are only
meant to be passed constant strings extracted from Angular application
templates, as passing other strings or variables could introduce XSS
vulnerabilities.
To better protect these APIs, turn them into template tags. This makes
it possible to assert that the associated template literals do not
contain any interpolation, and thus must be constant.
Also add tests for the change to prevent regression.
PR Close#40082
The linker is implemented using a Babel transform such that Babel needs
to parse and walk a source file to find the declarations that need to be
compiled. If it can be determined that a source file is known not to
contain any declarations the parsing and walking can be skipped as a
performance improvement. This commit adds an exposed function for tools
that integrate the linker to use to allow short-circuiting of the linker
transform.
PR Close#40137
The types of directives and pipes that are used in a component's
template may be emitted into the partial declaration wrapped inside a
closure, which is needed when the type is declared later in the module.
This poses a problem for JIT compilation of partial declarations, as
this closure is indistinguishable from a class reference itself. To mark
the forward reference function as such, this commit changes the partial
declaration codegen to emit a `forwardRef` invocation wrapped around
the closure, which ensures that the closure is properly tagged as a
forward reference. This allows the forward reference to be treated as
such during JIT compilation.
PR Close#40117
Currently when `ɵɵtemplate` and `ɵɵelement` instructions are generated by compiler, all static attributes are
duplicated for both instructions. As a part of this duplication, i18n translation blocks for static i18n attributes
are generated twice as well, causing duplicate entries in extracted translation files (when Ivy extraction mechanisms
are used). This commit fixes this issue by introducing a cache for i18n translation blocks (for static attributes
only).
Also this commit further aligns `ɵɵtemplate` and `ɵɵelement` instruction attributes, which should help implement
more effective attributes deduplication logic.
Closes#39942.
PR Close#40077
Durring analysis we find template parse errors. This commit changes
where the type checking context stores the parse errors. Previously, we
stored them on the AnalysisOutput this commit changes the errors to be
stored on the TemplateData (which is a property on the shim). That way,
the template parse errors can be grouped by template.
Previously, if a template had a parse error, we poisoned the module and
would not procede to find typecheck errors. This change does not poison
modules whose template have typecheck errors, so that ngtsc can emit
typecheck errors for templates with parse errors.
Additionally, all template diagnostics are produced in the same place.
This allows requesting just the template template diagnostics or just
other types of errors.
PR Close#40026
Refactors the i18n error tests to be unit tests in ngtsc_spec.ts. There
is two reasons for doing this.
First is that the tests in compliace_old expected an expection to be be
thrown but did not fail the test if no exception was thrown. That means
that this test could miss catching a bug. It is also a big hacky to call
compile directly and expect an exception to be thrown for diagnostics.
Also, this can easily be unit tested and an end-to-end test is not
necessary since we are not making use of the goldfiles for these tests.
It is easier to maintain and less hacky to validate that we get helpful
error messages when nesting i18n sections by calling getDiagnostics
directly.
PR Close#40026
This commit temporarily excludes classes declared in .d.ts files from checks
regarding whether providers are actually injectable.
Such classes used to be ignored (on accident) because the
`TypeScriptReflectionHost.getConstructorParameters()` method did not return
constructor parameters from d.ts files, mostly as an oversight. This was
recently fixed, but caused more providers to be exposed to this check, which
created a breakage in g3.
This commit temporarily fixes the breakage by continuing to exclude such
providers from the check, until g3 can be patched.
PR Close#40118
This commit replaces `bazel` with `yarn bazel` in the error message (that instructs to regenerate golden file)
thrown while executing compliance tests. We use `yarn bazel` in other places (so we use the local version of bazel,
not the global one).
PR Close#40078
Prior to this change, the `setClassMetadata` call would be invoked
inside of an IIFE that was marked as pure. This allows the call to be
tree-shaken away in production builds, as the `setClassMetadata` call
is only present to make the original class metadata available to the
testing infrastructure. The pure marker is problematic, though, as the
`setClassMetadata` call does in fact have the side-effect of assigning
the metadata into class properties. This has worked under the assumption
that only build optimization tools perform tree-shaking, however modern
bundlers are also able to elide calls that have been marked pure so this
assumption does no longer hold. Instead, an `ngDevMode` guard is used
which still allows the call to be elided but only by tooling that is
configured to consider `ngDevMode` as constant `false` value.
PR Close#39987
These tests started failing because they had type-check
errors in their templates, and a recent commit turned on
full template type-checking by default.\
This commit fixes those templates and updates the expected
files as necessary.
PR Close#40040
These tests do not pass the typecheck phase of the compiler and fail.
The option to disable typechecking was removed recently so these tests
need to be fixed to be valid applications.
PR Close#40033
A couple reasons to justify removing the flag:
* It adds code to the compiler that is only meant to support test cases
and not any production. We should avoid code in that's only
meant to support tests.
* The flag enables writing tests that do not mimic real-world behavior
because they allow invalid applications
PR Close#40013
The golden files for the partial compliance tests need to be updated
with individual Bazel run invocations, which is not very ergonomic when
a large number of golden files need to updated. This commit adds a
script to query the Bazel targets that update the goldens and then runs
those targets sequentially.
PR Close#39989
This test migrates source-mapping tests to the new compliance test framework.
The original tests are found in the file at:
`packages/compiler-cli/test/ngtsc/template_mapping_spec.ts`.
These new tests also check the mappings resulting from partial compilation
followed by linking, after flattening the pair of source-maps that each
process generates.
Note that there are some differences between the mappings for full compile
and linked compile modes, due to how TypeScript and Babel use source-span
information on AST nodes. To accommodate this, there are two expectation
files for most of these source files.
PR Close#39939
This commit allows compliance test-cases to be written that specify
source-map mappings between the source and generated code.
To check a mapping, add a `// SOURCE:` comment to the end of a line:
```
<generated code> // SOURCE: "<source-url>" <source code>
```
The generated code will still be checked, stripped of the `// SOURCE` comment,
as normal by the `expectEmit()` helper.
In addition, the source-map segments are checked to ensure that there is a
mapping from `<generated code>` to `<source code>` found in the file at
`<source-url>`.
Note:
* The source-url should be absolute, with the directory containing the
TEST_CASES.json file assumed to be `/`.
* Whitespace is important and will be included when comparing the segments.
* There is a single space character between each part of the line.
* Newlines within a mapping must be escaped since the mapping and comment
must all appear on a single line of this file.
PR Close#39939
Previously one could set a flag in a `TEST_CASES.json` file to exclude
the test-cases from being run if the input files were being compiled
partially and then linked.
There are also scenarios where one might want to exclude test-cases
from "full compile" mode test runs.
This commit changes the compliance test tooling to support a new
property `compilationModeFilter`, which is an array containing one or
more of `"full compile"` and `"linked compile"`. Only the tests
whose `compilationModeFilter` array contains the current compilation
mode will be run.
PR Close#39939
Previously files were serialized with an extra newline seperator that
was not removed when parsing. This caused the parsed file to start with
an extra newline that invalidated its source-map.
Also, the splitting was producing an empty entry at the start of the extracted
golden files which is now ignored.
PR Close#39939
The schema accidentally included the `expectedErrors` and `extraCheck`
properties below the `files` property instead of below the `expectations`
property.
PR Close#39939
The partial compiler will add a version number to the objects that are
generated so that the linker can select the appropriate partial linker
class to process the metadata.
Previously this version matching was a simple number check. Now
the partial compilation writes the current Angular compiler version
into the generated metadata, and semantic version ranges are used
to select the appropriate partial linker.
PR Close#39847
The newly built compliance test runner was not using the shared source
file cache that was added in b627f7f02e,
which offers a significant performance boost to the compliance test
targets.
PR Close#39956
Previously, if a component had an external template with a hard error, the
compiler would "forget" the link between that component and its NgModule.
Additionally, the NgModule would be marked as being in error, because the
template issue would prevent the compiler from registering the component
class as a component, so from the NgModule it would look like a declaration
of a non-directive/pipe class. As a combined result, the next incremental
step could fix the template error, but would not refresh diagnostics for the
NgModule, leading to an incrementality issue.
The various facets of this problem were fixed in prior commits. This commit
adds a test verifying the above case works now as expected.
PR Close#39923
Previously, if a trait's analysis step resulted in diagnostics, the trait
would be considered "errored" and no further operations, including register,
would be performed. Effectively, this meant that the compiler would pretend
the class in question was actually undecorated.
However, this behavior is problematic for several reasons:
1. It leads to inaccurate diagnostics being reported downstream.
For example, if a component is put into the error state, for example due to
a template error, the NgModule which declares the component would produce a
diagnostic claiming that the declaration is neither a directive nor a pipe.
This happened because the compiler wouldn't register() the component trait,
so the component would not be recorded as actually being a directive.
2. It can cause incorrect behavior on incremental builds.
This bug is more complex, but the general issue is that if the compiler
fails to associate a component and its module, then incremental builds will
not correctly re-analyze the module when the component's template changes.
Failing to register the component as such is one link in the larger chain of
issues that result in these kinds of issues.
3. It lumps together diagnostics produced during analysis and resolve steps.
This is not causing issues currently as the dependency graph ensures the
right classes are re-analyzed when needed, instead of showing stale
diagnostics. However, the dependency graph was not intended to serve this
role, and could potentially be optimized in ways that would break this
functionality.
This commit removes the concept of an "errored" trait entirely from the
trait system. Instead, analyzed and resolved traits have corresponding (and
separate) diagnostics, in addition to potentially `null` analysis results.
Analysis (but not resolution) diagnostics are carried forward during
incremental build operations. Compilation (emit) is only performed when
a trait reaches the resolved state with no diagnostics.
This change is functionally different than before as the `register` step is
now performed even in the presence of analysis errors, as long as analysis
results are also produced. This fixes problem 1 above, and is part of the
larger solution to problem 2.
PR Close#39923
