The browserUrlTree is only used to support the onSameUrlNavigation: 'ignore' logic. We can achieve this functionality without having this state tracked inside the Router. Instead, we can re-examine what ignore means: We don't want to rerun the matching logic, guards, or resolvers when we already know that nothing is changing.
Outside of the "navigated", there are two things that constitute a "change":
1. The browser URL might change. Because of skipLocationChange, the browser URL might not always match the internal state of the Router (we can navigate to a path but skip updating the browser URL). If we're navigating to a place that would change the browser URL, we should process the navigation. Theoretically, all we need to really do is update the browser URL instead of processing the whole navigation w/ guards, redirects, and resolvers. But this doesn't matter that much because the default value for runGuardsAndResolvers will skip all of this anyways.
2. The internal state of the Router might change. That is, we're navigating to a new path and may or may not be updating the updating the browser URL.
If either of the above are true, we process the navigation. If both are false, we aren't changing anything so we can safely ignore the navigation request (as long as onSameUrlNavigation === 'ignore').
Why is this change important?
* Simplification of Router internals. The Router has a lot of special case handling and one-offs to handle a limited set of scenarios. Removing these when possible makes the code easier to follow
PR Close#48065
Adds a utility to the `BoundTarget` that helps with resolving which element a deferred block is pointing to. We need a separate method for this, because deferred blocks have some special logic for where the trigger can be located.
PR Close#51816
When the `TargetBinder` was written, the only embedded-view-based nodes were templates, but now we have `{#if}`, `{#switch}` and `{#defer}` which have similar semantics. These changes rework the binder to account for the new nodes.
PR Close#51816
Content project allows the content to specify its own selector for matching against content projection slots, using the `ngProjectAs` special attribute. We can now treat this attribue specially, and generate the appropriate flag in the consts array, followed by the parsed CSS selector.
PR Close#51544
Supporting content projection requires us to emit three new kinds of output:
1. An `ngContentSelectors` field on the component metadata, which points to an array in the constant pool with all of the `select` attributes from `<ng-content>` elements.
2. One `projectionDef` instruction at the beginning of each root view template function for a component. That `projectionDef` points to a constant pool expression, which contains *parsed* selectors for all `<ng-content>` elements in the root's entire view tree.
3. A `projection` instruction for each `<ng-content>` slot in the view tree. These each get a data slot, a monotonically increasing "content slot", and a pointer to the tag's attributes in the component const array.
We support the first two features entirely within a new compilation phase.
The third feature, collection of processed attributes, is a bit trickier. We now treat `<ng-content>` tags as element-like ops, and use the normal attribute ingestion pipeline to process any attributes, and assign the appropriate `ConstIndex`.
**Note**: We also split up a number of the tests into two expectations files, one for the view functions, and one for other const listerals from the constant pool. This is because `TemplateDefinitionBuilder` emits the literals in a quirky order (mixed in with the view functions) due to how it lazily generates view functions. Our eager ordering is totally different, but by splitting the expectations, we can still share the same tests with `TemplateDefinitionBuilder`.
PR Close#51544
The new built-in control flow design includes calls to the `template` instruction with fewer arguments. This was previously handled implicitly, but it's more extensible to add an explicit flag to the template op to handle this case.
PR Close#51544
Although there are several pages, only `service-worker-getting-started.md` has an example.
That example code was migrated.
Worked around the few references to NgModule.
Other changes as follows:
**Revised "Getting Started"**
The original service worker example page (`service-worker-getting-started.md`) is full of holes! I don’t know how anyone got this working just by following this page
- it doesn’t tell you where to get the example (and it isn’t generated by AIO)
- it doesn’t tell you that ng serve doesn’t support service worker apps
- it mentions http-server in passing but not what that is or where to get it.
- several commands tell you to supply the “project-name” but (a) it doesn’t tell you how to find that name and (b) you never actually have to provide a name at all. That includes on the command that gets http-server going.
**Hid section on `ngsw-config` tool**
Hid the discussion of the `ngsw-config` tool in `service-worker-config.md` because
(a) the command invoke it didn't seem to work and (b) it referenced `<project-name>`
without explaining what that is or where to find it.
Left the hidden text in the page in case someone wants to do something about this.
Meanwhile, you seem to be able to edit the `ngsw-config.json` and then rebuild.
That should suffice.
**Disabled E2E**
The AIO CI will no longer attempt to run the (thin) E2E tests.
These tests could lightly test the app with the dev server but couldn't test the Service Worker aspects; that would require launching the app with something other than `ng serve` and the AIO CI is not set up for that.
The existing tests appeared to pass when run locally with the dev server.
However, only one of them actually ran; test runner reported.
"There were tests whose specified size is too big."
Decided to bypass E2E testing of this project for now.
PR Close#51687
The `malformedUriErrorHandler` is used as a recovery mechanism for when the `UrlSerializer`
throws an error when parsing a URL string. If custom error handling is
desired for this, it should instead be done inside the
`UrlSerializer.parse` method itself. There's no reason to have an entire
feature option built around what can otherwise just be `try...catch`.
BREAKING CHANGE: `malformedUriErrorHandler` is no longer available in
the `RouterModule.forRoot` options. URL parsing errors should instead be
handled in the `UrlSerializer.parse` method.
PR Close#51745
Currently deps tracker includes the exported scope of the exported NgModule only in the exported scope of that NgModule. This is in agreement with what AoT does today. But JIT diverges from this behavior by including these exported scopes into the compilation scope as well. Since deps tracker is going to be used for both AoT (local compilation mode) and JIT, the question might be which behavior the deps tracker should follow? Today it follows the AoT one, but it breaks some tests in Google which seem to depend on this behavior of JIT. So it is better to migrate deps tracker to what JIT does. This leads to a wider compilation scope in local compilation compared to full compilations, but it won't break any existing thing.
PR Close#51791
This change contains runtime logic needed to flatten the NgModule bootstrap field in local compilation mode. While it is quite odd to pass a "nested" array as NgModule bootstrap, it is still required to support this case in local compilation mode since it is supported in full compilation mode.
PR Close#51767
Today in local compilation mode the NgModule bootstrap definition is moved as it is into the runtime `ɵɵdefineNgModule`. This runtime was initially made for AoT full compilation mode and assumes that the bootstrap info is already flattened and resolved. This is not the case in local compilation where the bootstrap is the raw expression coming from the NgModule decorator and can be a nested array. To get around this problem we move the bootstrap along with other scope info (e.g., declarations, imports, exports) to the runtime`ɵɵsetNgModuleScope` to be further analyzed and flattened in runtime.
PR Close#51767
This change flattens the imports info on standalone component decorators in runtime dev mode by adding flattening logic to the deps tracker. Such flattening has no effect in AoT full compilation mode since these arrays are already resolved and flattened by AoT static analysis, but in local compilation mode it is needed since the raw array as appears on the component decorator will be passed to the deps tracker, and so it needs to be flattened.
This change does not affect prod runtime since deps tracker is only used in dev mode.
PR Close#51767
This change flattens the imports/exports/declarations info on ngModule decorators in runtime dev mode by adding flattening logic to the runtime `ɵɵsetNgModuleScope`. Such flattening has no effect in AoT full compilation mode since these arrays are already resolved and flattened by AoT static analysis, but in local compilation mode it is needed since the raw array as appears on the NgModule decorator will be passed to the runtime `ɵɵsetNgModuleScope`, and so it needs to be flattened.
This change has to effect on prod runtime as `ɵɵsetNgModuleScope` is not used in prod.
PR Close#51767
Currently when ESBuild bundles an application importing from
`@angular/core`, the signals library will be discovered during
export analysis. ESBuild will come across the constants for the reactive
signal graph- and end up considering some of these as side-effects given
the pattern of using a spread assignment for extending from e.g.
`REACTIVE_NODE` (a similar issue may occur if we e.g. extend from the
computed reactive node).
See more details on the issue: https://github.com/evanw/esbuild/issues/3392
Even though, ESBuild preserves these constants now, and all of its
dependencies— Terser will consider these as side-effect free and
eliminate these constants. This may require multiple passes though, and
might not be sufficient, depending on the chain of reactive node
extensions. E.g. in the signals branch we noticed some constants
unnecessarily being preserved.
PR Close#51809
Based on top of #51717
This commit adds extraction for enums, pipes, and NgModules. It also adds a couple of tests for JsDoc extraction that weren't covered in the previous commit.
PR Close#51733
Based on top of #51713
This commit adds docs extraction for information provided in JsDoc comments, including descriptions and Jsdoc tags.
PR Close#51733
Based on top of #51697
Adds extraction for accessors (getters/setters), rest params, and resolved type info for everything so far. This also refactors function extraction into a new class and splits tests for common class info and directive info into separate files.
PR Close#51733
Based on top of #51685
This expands on the extraction with information for directives, including inputs and outputs. As part of this change, I've refactored the extraction code related to class and to directives into their own extractor classes to more cleanly separate extraction logic based on type of statement.
PR Close#51733
Based on top of #51682
This expands on the skeleton previously added to extract docs info for classes, including properties, methods, and method parameters. Type information and Angular-specific info (e.g. inputs) will come in future PRs.
PR Close#51733
This commit adds a barebones skeleton for extracting information to be used for extracting info that can be used for API reference generation. Subsequent PRs will expand on this with increasingly real extraction. I started with @alxhub's #51615 and very slightly polished to get to this minimal commit.
PR Close#51733
This commit adds support for zone.js 0.14.0 and drops support for older versions
BREAKING CHANGE: Angular now required `zone.js` version `~0.14.0`
PR Close#51774
Another try at deflaking the tests on Windows. I'm trying a couple of fixes here:
1. I noticed that it's usually the indexer tests that fail during flaky runs. These tests also happen to be the only ones that don't pass in the `files` argument of `NgtscTestEnvironment.setup`. When `files` isn't passed in, we don't hit the file path that sets up the `MockFileSystem`. With these changes I make it so that we always initialize the mock file system.
2. The missing file system error usually comes from the `absoluteFrom` call that initializes the optional `workingDir` argument. My theory is that because it's a default value for an argument, it gets called too early before everything is initialized. These changes move the `absoluteFrom` call further down until it's needed.
PR Close#51804
upgrade the warning for lazy-loaded lcp images when using NgOptimizedImage to an error
BREAKING CHANGE:
Previously when NgOptimizedImage directive detected that an LCP image is lazy-loaded, a console warning was produced. Now the directive throws an error to make it more discoverable in a console. If you receive this error, refer to this guide for additional information: https://angular.io/guide/image-directive#step-4-mark-images-as-priority
PR Close#51748
Terser does not treat `Symbol` as side-effect free- so if we end up with
a symbol export being loaded, it will result in the symbol being
retained.
We noticed this in the signals prototyping where symbols exported
from `computed` ended up appearing in symbol bundling tests.
PR Close#51776
This commit moves the destroy logic from 'effect' in the lower-level
'watch' so this implementation is shared among varius watch implementations.
PR Close#51757
Certain code patterns and tools in Google (and possibly 3P world) lead to the situation that a component is bootstrapped/rendered without its ng-module being loaded in the browser. Technically speaking this should be an anti-pattern since the ng-module could contain some runtime logic (e.g., providing something, calling some services, etc) and its not being loaded leads to unexpected behaviour. However, in many cases ng-module is an empty class and its only usage is for providing scope, and since in AoT full compilation mode we already hard-code dependencies into components so we can get away with not loading the ng-module. But in AoT local compilation mode it is not possible to get away since the component's dependencies are computed in runtime and the presence of the corresponding ng-module in the browser is needed. For this reason in this change it is forbidden to attempt to render a component without first loading its ng-module in local compilation mode and an explicit error message is created to make this situation clear. This error message can help with catching such cases when running TGP in Google.
It would be an interesting question as to whether to ban this situation in full compilation mode as well, as it is error prone and these errors are sometimes very hard to debug.
PR Close#51726
