Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51653
Changelog: [Internal]
Adds a transform that ensures no types and interfaces named `Props` end up in the generated TypeScript definitions. Those are not descriptive and cause duplicate types in the rollup.
Reviewed By: huntie
Differential Revision: D75508800
fbshipit-source-id: 8d64ec19cbabe57495d6462df9d372ba42cda618
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51797
Changelog: [internal]
Fantom tests are generally very reliable, but a source of flakiness seems to be coming from `result.stderr` being `undefined` and throwing in this callsite. We just add a check to work around it, hoping it's just being set to `undefined` because there really are no errors.
Reviewed By: rshest
Differential Revision: D75942405
fbshipit-source-id: d8920cc2573f029de6bb108310c31f1992e74301
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51774
Changelog: [internal]
Just a small refactor to remove some unnecessary `using namespace` directives for code in the `facebook::react` namespace.
Reviewed By: javache
Differential Revision: D75874213
fbshipit-source-id: e63a91b412b2865eb05961d2174b14230ff35e88
Summary:
This is similar to https://github.com/facebook/react-native/issues/51590, but way better as it improves indirectly the flakiness for tests in the API tab. When the logbox is shown in debug mode, it interferes and sometimes makes that test fail, so this prevents that.
Android also takes more advantage of the improvement with this change, the previous PR only improved significantly iOS.
All the screens inside the RNTester seem to have a deeplink, which makes it easier to open the tests as the test cases are intended to check mostly specific behaviour of RN, and it is not necessary to have a middle step to find the specific components.
Maybe it would be good to run this a few times in CI to see if there are no side effects or flakiness added by opening deep links on CI builds.
## Changelog:
[INTERNAL] - Improve e2e times by using deep links to open examples
Pull Request resolved: https://github.com/facebook/react-native/pull/51786
Test Plan:
```sh
yarn e2e-test-android
yarn e2e-test-ios
```
iOS:
| Before | After |
|--------|-------|
| <img width="387" alt="image" src="https://github.com/user-attachments/assets/03ccd957-d401-4944-bb5c-d3e7db957b2e" /> | <img width="364" alt="image" src="https://github.com/user-attachments/assets/40a14c95-63f8-441d-b718-b5f57a506393" /> |
Android:
| Before | After |
|--------|-------|
| <img width="455" alt="image" src="https://github.com/user-attachments/assets/c71da8d0-df69-44af-b1b2-580995ce55c7" /> | <img width="449" alt="image" src="https://github.com/user-attachments/assets/7357e670-3510-4bbe-8543-68d3bd8c4bea" /> |
Reviewed By: cipolleschi
Differential Revision: D75938844
Pulled By: cortinico
fbshipit-source-id: c7d4063af561e7b0e583eddefcbb289786f3805a
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51776
changelog: [internal]
testing animation must be predictable.
When animation takes 1 second and test calls `unstable_advanceAnimationsByTime(1000)`, the expectation is that the animation would have completed.
Previously it was necessary to wait for at least 38ms over the time it took to complete animation. This diff reduces the delta to 22ms.
Reviewed By: andrewdacenko
Differential Revision: D75813087
fbshipit-source-id: 1a27e5c227a0f3f1183ebdf41dc8c18149d68be0
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51757
These both have a boolean flag `hasBeenMeasured_` which is never set to true, so the cached measurement is never used, but it would be wrong if it was used, since the cache doesn't respect the content, or measure constraints. Delete the broken code.
Changelog: [Internal]
Reviewed By: alanleedev
Differential Revision: D75817999
fbshipit-source-id: eb062f523840bbac7099b28f7e9b25cf9ad302aa
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51773
Changelog: [internal]
(This isn't a public API yet so not marking it as a bugfix).
This fixes a crash in the `offsetParent` DOM API when called on a node whose parent is the `documentElement`.
Reviewed By: andrewdacenko
Differential Revision: D75876349
fbshipit-source-id: d90fcc3cb98f592c396cb0f5ee6ec80c9570745a
Summary:
This PR makes sure React Native still builds with USE_HERMES=0.
## Changelog:
[IOS] [FIXED] - make RNTester build with USE_HERMES=0
Pull Request resolved: https://github.com/facebook/react-native/pull/51772
Test Plan: CI Green
Reviewed By: andrewdacenko
Differential Revision: D75876484
Pulled By: cipolleschi
fbshipit-source-id: 72f3a7d33c19063b52c4c3e577ad2c5eae049d90
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51771
This made build_android fail because the parameter name mismatched between class and one of the implementation.
Changelog:
[Internal] [Changed] -
Reviewed By: cipolleschi
Differential Revision: D75869827
fbshipit-source-id: a40dd1fb914e0694f1958c8b19eb9672de683c75
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51748
Changelog: [Internal]
Implements the first RNDT shell-specific feature based on https://github.com/facebook/react-native-devtools-frontend/pull/168 - namely, the ability for RNDT to foreground itself when certain events occur. This is most noticeable when pausing on a breakpoint.
Reviewed By: huntie, vzaidman
Differential Revision: D75795689
fbshipit-source-id: a073bf8ea96ba70d835007f5af6069d49a693d81
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51767
This is going to convert the last class inside `com.facebook.react.views.view` to Kotlin.
I still need to assess if this is going to be breaking or not for our users. Potentially some
breakages for Kotlin consumer could happen (see changes on ReactViewManager).
I'll follow-up with an investigation on that one.
Changelog:
[Android] [Changed] - Convert ReactViewGroup to Kotlin
Reviewed By: mdvacca
Differential Revision: D75797215
fbshipit-source-id: 5d58faa651d5959909a4b0237aad9c4209895083
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51752
changelog: [internal]
Introduce a way to test animations: `unstable_advanceAnimationsByTime`. An API that fakes passage of time and triggers UI ticks to simulate how animations are run on iOS and Android.
The API is marked as unstable because it might evolve as we write more tests for C++ Animated.
Reviewed By: mdvacca
Differential Revision: D75787082
fbshipit-source-id: 24e29732bbbf581871e7868289257ab60d891ddf
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51751
changelog: [internal]
these asserts are not correct with C++ Animated or anything that overrides pull transaction. When pull transaction is overriden, the modified mounting instructions are not stored in shadow tree. Therefore, these asserts fail.
Reviewed By: rshest
Differential Revision: D75787083
fbshipit-source-id: ac08b96f6dea89a9f127b179a5d05e351b281c35
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51750
changelog: [internal]
inline call `onRender()` instead of calling it from `startRenderCallbackIfNeeded`. This has identical functionality but allows to use Fantom to test C++ Animated. In Fantom, there is only one thread the existing mechanism in C++ Animated uses thread locals to capture which thread in the UI thread. Therefore, some assumptions are broken. This is just an easy workaround around the problem.
Reviewed By: javache, zeyap
Differential Revision: D75787084
fbshipit-source-id: 8d70ef2d13e4ab14807deb7ca4cb0844980f5901
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51749
changelog: [internal]
Make it possible to inject time via `now` argument to C++ Animated. This will be used in testing.
Reviewed By: javache
Differential Revision: D75710463
fbshipit-source-id: 2d6da875c7379c4b229f8b7af0fa665cebc2ca8b
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51726
Previously we were spreading the searchPath on the FBLazyVectorPath everywhere, but that's not needed.
This change ensure that only the targets that nedds visibility over the FBLazyVectorPath depends on them.
## Changelog:
[Internal] -
Reviewed By: cortinico
Differential Revision: D75756304
fbshipit-source-id: 41abe073bf271fd5059f9cb8536cc4a8948a03dd
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51727
Previously we were spreading the searchPath on the ReactFBReactNativeSpecPath everywhere, but that's not needed.
This change ensure that only the targets that nedds visibility over the ReactFBReactNativeSpecPath depends on them.
## Changelog:
[Internal] -
Reviewed By: cortinico
Differential Revision: D75756302
fbshipit-source-id: 41680b98dd07ea1abb09cebb1a1b8321f08929a0
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51728
Previously we were spreading the searchPath on the CallInvoker everywhere, but that's not needed.
This change ensure that only the targets that nedds visibility over the CallInvoker depends on them.
## Changelog:
[Internal] -
Reviewed By: cortinico
Differential Revision: D75756303
fbshipit-source-id: d6fe70e43f65d8e9da5f1d18dc58491217575734
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51729
Previously we were spreading the searchPAth on the Runtime Executor everywhere, but that's not needed.
This change ensure that only the targets that nedds visibility over the runtimeExecutor depends on them.
## Changelog:
[Internal] -
Reviewed By: cortinico
Differential Revision: D75756305
fbshipit-source-id: e54f1939d72049e7f4e8d3d14f7a33eacd5ac46f
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51730
This change cleans up the header search paths that are now mostly computed starting from other informations of the target.
There are still some execptions, for example for header only target for which we can't have a SwiftPM target. Therefore targets that depends on header only targets need to specify them in their search path.
Plus there are a small number of other exceptions, such as some circular dependencies on header only files.
We will fix them later on, by introducing separated targets
## Changelog:
[Internal] -
Reviewed By: cortinico
Differential Revision: D75756307
fbshipit-source-id: 53f24dcb08947648c8fce5c520a949b3b896c2fa
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51731
With this change, we are computing the headerSearchPaths starting from the path, and the source's path. In this way we can greatly reduce the amount of searchPaths we have to manually specify, making it much easier to maintain the SwiftPM package in the long run.
## Changelog:
[Internal] -
## Facebook:
The next diff in the stack proceeds with the cleanup.
Reviewed By: cortinico
Differential Revision: D75756306
fbshipit-source-id: 411ba6b598e568648488921c3ee62b8111f76749
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51691
In cocoapods infrastructure, Scrollview is part of ReactFabric.
In the SwiftPM implementation we erroneously moved that to reactFabricComponents. However, the new view culling features in reactFabric imports scrollview and that was broken in Swift PM.
This change fixes it.
## Changelog:
[Internal] - move scrollview to reactFabric
Reviewed By: lenaic
Differential Revision: D75638510
fbshipit-source-id: 00828956093aaa87e15c22d15ca3a647876182f5
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51760
This API is evil.
Yoga's public API never allows a dirty node to become clean again, until its laid out, but this API requires doing that, since we will otherwise automatically dirty by default.
Let's replace it with `YogaLayoutableShadowNode::shouldNewRevisionDirtyMeasurement()`, which lets individual ShadowNodes represent whether a new revision's state and props should cause dirtying, defaulting to true.
Changelog:
[General][Removed] - Remove `YogaLayoutableShadowNode::cleanLayout()`
Reviewed By: lenaic
Differential Revision: D75479902
fbshipit-source-id: a40aa531522a76dc49feb7d12bae5a8d877c6c06
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51719
When `Animated` traverses props for instances of `AnimatedNode`, it flattens `props.style` before traversing it so that we correctly ignore any `AnimatedNode` instances that may be shadowed by static values:
```
{
style: [
{transform: [{translateX: new Animated.Value(0)}]},
{transform: [{translateX: 100}]},
],
}
```
However, there is a bug that occurs when *every* `AnimatedNode` instance is shadowed. In this case, `AnimatedProps` assumes that there are *no* `AnimatedNode` instances in the entire `props.style`.
It then incorrectly operates on the unflattened `props.style`, which *does* have `AnimatedNode` instances. When this is passed to `View`, the `AnimatedNode` instances are encountered and can cause a crash in `processTransform`, as reported by: https://github.com/facebook/react-native/issues/51395
The fix for this was originally attempted by riteshshukla04 in https://github.com/facebook/react-native/pull/51442. This diff reuses the same unit test case, but it applies a different fix that does not involve re-traversing the `props.style` object.
The fix is gated behind a feature flag, `alwaysFlattenAnimatedStyles`. This will enable us to validate correctness of the new behavior before enabling it for everyone. (Beyond fixing the bug described above, this also causes styles to flatten more aggressively, so production testing is important to ensure stability.)
Changelog:
[General][Changed] - Creates a feature flag that changes Animated to no longer produce invalid `props.style` if every `AnimatedNode` instance is shadowed via style flattening.
Reviewed By: javache
Differential Revision: D75723284
fbshipit-source-id: 504f63e8edf836243d615783e119137a920ad271
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51718
## what
the current removing listeners code is not working. in prod, when removeScrollListener is called, it creates a new weakReference of the scroll listener and it would never match any existing weakreference listeners in the list. The remove action would end up removing nothing from the list.
This diff fixes the issue by iterating through the list and compare the obj
reference: https://stackoverflow.com/questions/6296051/how-to-remove-a-weakreference-from-a-list
Changelog: [Internal]
Reviewed By: Abbondanzo
Differential Revision: D75716993
fbshipit-source-id: af9eed218deb44e87317c19ffca7d227b96de8e2
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51716
Adds a feature flag to experiment with different `prerenderRatio` values from the `VirtualView` native component. Notably, this is the first time that the a feature flag has a non-boolean type.
Changelog:
[Internal]
Reviewed By: mdvacca
Differential Revision: D75713594
fbshipit-source-id: ff3feb8d8bb3292fea6e04a4cfccd87e8249144e
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51713
Changes the React Native Feature Flags system so that numeric feature flags are represented in native languages as double instead of int, in order to avoid loss of precision for non-integral JavaScript numbers.
Changelog:
[Internal]
Reviewed By: mdvacca
Differential Revision: D75709111
fbshipit-source-id: d4b2d553ce1e719a5f4b136a4f8d2e46446ef8d8
Summary:
Following up the announcement made at AppJS, this change stops testing the legacy architecture in our CI
## Changelog:
[Internal] - Stop testing the legacy architecture in CI
Pull Request resolved: https://github.com/facebook/react-native/pull/51738
Test Plan: waiting for GHA
Reviewed By: cortinico
Differential Revision: D75791359
Pulled By: cipolleschi
fbshipit-source-id: cb3159338835f49589fa6f495cfb9f47750825fe
Summary:
I'm writing some tests for some APIs/components, and I see some opportunity for improvement in the way how we can write e2e cases for the RNTester.
This diff improves two things:
1. e2e execution times: right now as we are using the `scrollUntilVisible` functionality, it takes quite some time *for some cases*, as some of the items are not very up in the lists and getting to them it's not always very fast.
2. Flakiness: I ran the tests multiple times locally, and in multiple occasions, the `scrollUntilVisible ` did not find anything as the scroll was too fast so the test ended up failing.
Instead of using `scrollUntilVisible` for all cases, we can simply use the search bar which we have in both Components and APIs tabs. This runs faster and we can also share the search flow across multiple test cases, so writing the tests becomes a bit simpler as well.
Initially, I did this for all cases but not all cases benefit from this change as some of them are easier to find than others – the improvement was most notable in iOS (keyboard visibility seems to make a big difference), but I still think it is a good baseline to use the search as if more test cases are added, likely, many of them are not going to be so easy to find.
## Changelog:
[INTERNAL] - Improve e2e times and flakiness by using the search
Pull Request resolved: https://github.com/facebook/react-native/pull/51590
Test Plan:
These are the differences in the time it takes to pass the tests locally.
iOS:
| Before | After |
|--------|-------|
| <img width="336" alt="image" src="https://github.com/user-attachments/assets/6bc80de6-5a10-4c0c-b6b0-f7850d746781" /> | <img width="329" alt="image" src="https://github.com/user-attachments/assets/de47a7ff-84cf-4916-abce-9b3df96a03e1" /> |
<details>
<summary>Android (no significant difference):</summary>
| Before | After |
|--------|-------|
| <img width="328" alt="image" src="https://github.com/user-attachments/assets/73023669-af8e-410c-aee2-529d1cce7acf" /> | <img width="336" alt="image" src="https://github.com/user-attachments/assets/3fb868a9-36f5-43c3-9261-6f2753e28e5f" /> |
</details>
Reviewed By: cortinico
Differential Revision: D75389138
Pulled By: cipolleschi
fbshipit-source-id: cf9d11ab0f84c91eaa20ee5c4441766729925571
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51746
I observed that when continue-on-error is set to true, Github reports the outcome of a job as success even if it fails.
With this change, we should ensure that when an E2E test fails, the pipeline fails, so that we can retry the jobs properly.
## Changelog:
[Internal] - Remove continue-on-error from e2e tests in GHA
Reviewed By: cortinico
Differential Revision: D75796284
fbshipit-source-id: 0e769f53d7355ae6c985aace334b23205780673a
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51745
Changelog: [Internal]
1. Updates debugger-frontend's sync-and-build script to include a simple Markdown changelog inline and -mention authors who are Meta employees.
2. Also adds a `--no-build` flag that is mainly helpful for iterating on changelog generation logic.
Reviewed By: hoxyq
Differential Revision: D75789680
fbshipit-source-id: b30b49a9c50e93e7161a2dad012b92ef124a3a16
Summary:
- This is similar to https://github.com/facebook/react-native/pull/43566, but include also updated `StackFrame` type.
The current `Devtools.d.ts` doesn't match the `parseErrorStack.js` and `symbolicateStackTrace.js` implementations.
I've tried to run `build-types`, but only `symbolicateStackTrace.d.ts` was generated. I've not checked why.
## Changelog:
<!-- Help reviewers and the release process by writing your own changelog entry.
Pick one each for the category and type tags:
[ANDROID|GENERAL|IOS|INTERNAL] [BREAKING|ADDED|CHANGED|DEPRECATED|REMOVED|FIXED|SECURITY] - Message
For more details, see:
https://reactnative.dev/contributing/changelogs-in-pull-requests
-->
[GENERAL] [FIXED] - Devtools TS Types
Pull Request resolved: https://github.com/facebook/react-native/pull/51737
Test Plan: Call `parseErrorStack` and `symbolicateStackTrace` in TS.
Reviewed By: rshest
Differential Revision: D75782013
Pulled By: huntie
fbshipit-source-id: 91fe560f079731af2a5834c8de8eafb723d00bf9
Summary:
## Changelog:
[General] [Changed] - Remove native animation fabric sync in JS and infinite animation loop workaround, when cxxNativeAnimated is enabled
when cxxNativeAnimated is enabled, we'll sync native animation props back to Fabric in native, when that happens we can remove the code in JS for same purpose
Reviewed By: sammy-SC
Differential Revision: D75789100
fbshipit-source-id: 966e2b187f43e8743ccbf7ba97b8e8a27273fe0c
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51743
Noticed some of these were missing, which may lose us useful test signal, or prevent tests from being ran at all.
Changelog: [Internal]
Reviewed By: sammy-SC
Differential Revision: D75789456
fbshipit-source-id: 75645866c672c77d3dac34383105955ef6d25e60
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/50317
`rncore`, `FBReactNativeSpec` and `FBReactNativeComponentSpec` contain the same symbols, which leads to conflicts when we try to merge them into a single shared library. Cleanup the duplication and standardize on `FBReactNativeSpec` everywhere. I've left the Android OSS targets names as is, to avoid breaking deps.
This aligns react-native's package.json with the codegen tooling supported across iOS and Android, which is a single target for all all type-derived codegen.
Changelog: [Internal]
Reviewed By: sammy-SC
Differential Revision: D55037569
fbshipit-source-id: dbf3c0a427c9d0df96e439b04e5b123cd1069c51
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51708
Changes `yarn featureflags` to default to `--update` when no options are supplied.
Provides a way to display the available options with a new `--help` option.
Changelog:
[Internal]
Reviewed By: sammy-SC
Differential Revision: D75692119
fbshipit-source-id: 50d25c5f12c7646159872661c93ebe8c42f44788
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/51707
Currently, the feature flag system does not properly support non-boolean feature flags. The type definitions and generated code make assumptions about the value type being `boolean`.
This diff fixes these assumptions so that non-boolean feature flags are supported.
Changelog:
[Internal]
Reviewed By: sammy-SC
Differential Revision: D75690997
fbshipit-source-id: 870063ff979d0650ce6d0a2f6b340d97c28d7d0b