Summary:
This diff fixes an edge case where scroll to the end of the list may trigger a bounce back effect.
This issue is a regression from D32487846 (https://github.com/facebook/react-native/commit/f70018b37532622f08f20b2c51cdbfca55d730ea) (See [this comment](https://www.internalfb.com/diff/D32487846 (https://github.com/facebook/react-native/commit/f70018b37532622f08f20b2c51cdbfca55d730ea)?dst_version_fbid=263960175698224&transaction_fbid=566201141113715)) that zero velocity fling at the end of the scroll view makes the next fling animator use previous post animation position. This is due to cached `postAnimationValue` is applied mistakenly.
- Pass velocity instead of velocity sign to the helper class
- Update helper class logic to decide if we need to use post animated value from last fling animation
Changelog:
[Internal]
Reviewed By: javache
Differential Revision: D32566010
fbshipit-source-id: 1c61659030151f8f2c7648ca901b8b4158835538
Summary:
This diff fixes an edge case where fling and snap failed to find the correct target position when children views not fill up all the scrollable space. In this case, the target position would be calculated as the end of the scrollable space, which case the snap logic to go to the end of the scrollable area, instead of stop at the expected snapping position.
Changelog:
[Android][Fixed] - Fix fling and snap with recycler viewgroup where fling to the end of scrollable distance when it goes over current rendered children views.
Reviewed By: mdvacca
Differential Revision: D32565459
fbshipit-source-id: 319ef6e2d4e1c4deb9e45ed02c1bff7d807575c3
Summary: Changelog: [Internal] Add a `isTaggedVersion` function to filter out commits from release automation.
Reviewed By: sota000
Differential Revision: D32842035
fbshipit-source-id: 14bb262a1d2a96ffda87c759a3202c4f9a356141
Summary:
Bump deprecated-react-native-prop-types dependency to ^2.3.0. This version contains a .windows.js fork of some deprecated props. Current version has missing information for Windows, breaking CI for react-native-windows repository.
## Changelog
[General] [Changed] - Upgrade deprecated-react-native-prop-types dependency
Pull Request resolved: https://github.com/facebook/react-native/pull/32700
Reviewed By: lyahdav, rozele
Differential Revision: D32816801
Pulled By: yungsters
fbshipit-source-id: f6a8507637075b1a936d91c264bbd8a462505f48
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/32698
Changelog: [Internal] - Renaming the eslint-plugin-codegen to eslint-plugin-specs as a better descriptor for this package. Part of the plan is to publish as a public npm package
Reviewed By: RSNara
Differential Revision: D32777702
fbshipit-source-id: 7776c481871031dba1280d87fb76b777ecc23339
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/32693
Changelog: [Internal] this diff adds dependecies to the buck files so that an internal app that referes RNtester will work as intended.
Reviewed By: cortinico
Differential Revision: D32778338
fbshipit-source-id: d4703903c2227283fe9c70f02bb7da420d936fbb
Summary:
changelog: [internal]
Background executor has been shipped on both platforms for a long time.
I've kept the flag around because I wanted to run tests and compare Concurrent Mode vs Background Executor. The intention was to see if we can get rid of Background Executor to simplify the threading model.
Since then, React team has moved away from Concurrent Mode towards more gradual rollout of concurrent rendering and it no longer makes sense to do this comparison. Right now, we don't have a concern with concurrent rendering and Background Executor. If we ever want to run the an experiment, this gating will need to be added again.
Reviewed By: javache
Differential Revision: D32674798
fbshipit-source-id: a1e51c9c5b8e48efa4cb0f25379d58e7eb80ccd9
Summary: Changelog: [Internal] Adding an option in RN Tester to enable the new codegen discovery option where it uses generate-artifacts.js to codegen native files. It also updates generate-artifacts.js to support the case where react-native is not within node_modules. It also updates the option name for rn-demo-app.
Reviewed By: cortinico
Differential Revision: D32777912
fbshipit-source-id: f2b76fa61573e3d4507a9f16f8243ac7ca006900
Summary:
I've noticed that the `performance.now()` implementations differ on iOS and Android.
iOS:
```objc
PerformanceNow iosPerformanceNowBinder = []() {
// CACurrentMediaTime() returns the current absolute time, in seconds
return CACurrentMediaTime() * 1000;
};
```
Android:
```c++
double reactAndroidNativePerformanceNowHook() {
auto time = std::chrono::steady_clock::now();
auto duration = std::chrono::duration_cast<std::chrono::nanoseconds>(
time.time_since_epoch())
.count();
constexpr double NANOSECONDS_IN_MILLISECOND = 1000000.0;
return duration / NANOSECONDS_IN_MILLISECOND;
}
```
For consistency, I thought why not just use the same implementation on both iOS and Android.
It also seems more logical to use Chrono on iOS, since it has nanosecond precision and we just multiply it to milliseconds, whereas `CACurrentMediaTime` multiplies to seconds, and we divide it down to milliseconds again.
## Changelog
(internal change only)
Pull Request resolved: https://github.com/facebook/react-native/pull/32695
Test Plan:
Run on iOS and Android:
```ts
const now = global.performance.now()
console.log(`${Platform.OS}: ${now}`)
```
Reviewed By: feedthejim
Differential Revision: D32793838
Pulled By: ShikaSD
fbshipit-source-id: e7967780be95956a75a3a3757311af0077976d23
Summary:
Changelog: [Internal] Remove warning about creating a pull request to a release branch
We want to support this use case as we want pick requesters to land the cherry-picks themselves and rely on CircleCI automation to test
Reviewed By: cortinico
Differential Revision: D32780819
fbshipit-source-id: 882625a016653dbf480da33fd499d3718b2b5524
Summary:
changelog: [internal]
UIManager::animationTick is thread safe. Let's make it obvious by marking the method const.
Reviewed By: javache
Differential Revision: D32669102
fbshipit-source-id: 49e35d0f0a5c5d1b03baa1cbf9cdece082909e85
Summary:
The current implementation of the prelive composer is setting the RN override preference to nil when live shopping ends. However, if the user had a previously set appearance override, this would clear it out.
To resolve this, I cache the user's override preference before we override it, and restore it when live shopping ends.
Additionally, I'm adding a missing callback to restore the override pref when live shopping is dismissed, but not cancelled.
Changelog:
[ReactiveNative][Appearance] - Add function to retrieve overridden RN appearance
Differential Revision: D32564739
fbshipit-source-id: d0b4e06d2bbadad68f172d951609cba3e4587e5d
Summary:
The current implementation of `AccessibilityInfo.announceForAccessibility` will immediately interrupt any existing in progress speech with the announcement. Sometimes this is desirable behaviour, but often you will want to wait until existing speech is finished before reading the new announcement. This change gives us that option.
My personal use case for this feature is a custom text input. When typing on iOS with voiceover enabled, each character is read out after being selected. I wanted to add some additional information after each character to help with the context of what has changed in the input, but I didn't want to override the reading of the character itself.
This feature is supported natively on iOS by constructing an `NSAttributedString` with the property [`accessibilitySpeechQueueAnnouncement`](https://developer.apple.com/documentation/foundation/nsattributedstring/key/2865770-accessibilityspeechqueueannounce), so this change just adds an extra parameter to `AccessibilityInfo.announceForAccessibility` which controls the value of that property on the native side. Adding this as an extra optional parameter with false as the default ensures that existing uses of the function won't be affected.
Unfortunately, this feature doesn't appear to be supported on Android, so the new second property will be iOS only.
## Changelog
[iOS] [Added] - add new argument to announceForAccessibility to allow queueing on iOS
Pull Request resolved: https://github.com/facebook/react-native/pull/32637
Test Plan:
I've updated the `announceForAccessibility` section in RNTester with multiple buttons to demonstrate the difference between `queue: false` (default) and `queue: true` and show they work as intended.
Here's the expectation for each button:
- "Announce for Accessibility Immediately": on press, should start reading the button label, then be interrupted by the announcement
- "Announce for Accessibility Queued": on press, should read the button label then read the announcement afterwards
- "Announce for Accessibility Queue Multiple": on press, should read the button label, then read three announcements sequentially, no interruptions
You can see the realisation of those expectations in the following video recorded on an iPhone 12 running iOS 15.0.2:
https://user-images.githubusercontent.com/14826539/142770536-d57bfd69-eba5-444d-9c89-4bf4851ea062.mov
I've also tested the same way on an iPhone 8 running iOS 13.4 and it works exactly the same.
Reviewed By: yungsters
Differential Revision: D32637989
Pulled By: philIip
fbshipit-source-id: 3e90add523f11eb0eb34ea623211249263f257e2
Summary:
I've been seeing a couple crashes related to missing hermes-executor-common.so, seems to happen on specific android versions, but can't repro. I investigated this so file more and noticed it is incorrectly linked as a static library here https://github.com/facebook/react-native/blob/b8f415eb6cdc0e0e7a7413b6f9defdcee304d9e8/ReactAndroid/src/main/java/com/facebook/hermes/reactexecutor/Android.mk#L20. There doesn't seem to be any reason for this to be a shared lib so I changed it to be compiled as a static lib.
## Changelog
[Android] [Fixed] - Make hermes-executor-common a static lib
Pull Request resolved: https://github.com/facebook/react-native/pull/32683
Test Plan:
- Verify there is no more hermes-executor-common-{release,debug}.so
- Test locally in an app to make sure it build and run properly.
- Verify that the crash happening on play store pre-launch report doesn't happen anymore.
Reviewed By: ShikaSD
Differential Revision: D32754968
Pulled By: cortinico
fbshipit-source-id: cb57e2d81edb4cbdb1f003dab45c53e594a5a62a
Summary:
Changelog: [Internal] Revert --sourcemap-output argument
Reverting on main from discussion from the 67 release:
https://github.com/reactwg/react-native-releases/discussions/1#discussioncomment-1705266
Original commit changeset: 0c2d98746b36
Reviewed By: GijsWeterings
Differential Revision: D32734743
fbshipit-source-id: cc3130581d01bdcdc9f5d483161ae9506fdbf7b8
Summary:
changelog: [internal]
Attempt at fixing a crash in `Scheduler::uiManagerDidCloneShadowNode` by setting delegate to nullptr explicitly.
I don't think this will make a difference because `scheduler_` is released at the end of `dealloc` but it is worth a shot. Maybe some complex interaction between Obj-C and C++ comes into play here.
In this diff I also removed `_animationDriver = nullptr` because dealloc will do that automatically.
Reviewed By: philIip
Differential Revision: D32720901
fbshipit-source-id: 227ced2331384c47e8d15a323ee8a621bbb3d179
Summary:
react-native-windows currently needs to maintain a fork of TextNativeComponent to wire through a native-only prop for `isPressable`.
The reason we do this on Windows is that we implement an optimization so we only attempt to hit test a virtual Text node if it is actually pressable, leading to significant perf improvement for pointer events (e.g., onMouseEnter / onMouseLeave) on Text.
Changelog:
[General][Added] - Native-only prop to optimize text hit testing on some RN platforms
Reviewed By: JoshuaGross
Differential Revision: D32564637
fbshipit-source-id: bf47c68d94a930d2c620cb3b1584355c5e412bd4
Summary:
Changelog: [Internal] Update CircleCI to auto-deploy release branch on push
This work is part of an effort to automate the release process by using a push to a release branch as a trigger to prepare, package and deploy react-native to npm from CircleCI
The following diagram describes the context (what kind of releases we do, relevant scripts and what they do), the pre-existing process for the different types of release and how I've modified the process.
{F683387103}
This diff updates the relevant CircleCI workflows
Reviewed By: sota000
Differential Revision: D32702420
fbshipit-source-id: e20cdeb53eb4a8ce7e54e083e3e14bd89e11b789
Summary:
Changelog: [Internal] - Extract logic from bump-oss-version specific to prod releases
This work is part of an effort to automate the release process by using a push to a release branch as a trigger to prepare, package and deploy react-native to npm from CircleCI
The following diagram describes the context (what kind of releases we do, relevant scripts and what they do), the pre-existing process for the different types of release and how I've modified the process.
{F683387103}
This diff creates the `prepare-package-for-release` script referenced by extracting it out of `bump-oss-version` and leveraging `set-rn-version`. It adds some helper functions to `version-utils` with tests
Reviewed By: sota000
Differential Revision: D32556610
fbshipit-source-id: eb4ddc787498744156f985ab6d205c5d160e279b
Summary:
Changelog: [Internal] Copy over universal (across dry-run, nightly, release) work in `bump-oss-version` script
This work is part of an effort to automate the release process by using a push to a release branch as a trigger to prepare, package and deploy react-native to npm from CircleCI
The following diagram describes the context (what kind of releases we do, relevant scripts and what they do), the pre-existing process for the different types of release and how I've modified the process.
{F683387103}
This diff creates the `set-rn-version` script referenced by extracting it out of `bump-oss-version`
Reviewed By: sota000
Differential Revision: D32556608
fbshipit-source-id: 6c2868c01ddd930375279a5105bcd0d447f65734
Summary:
Changelog: [Internal] - Add getNextVersionFromTags to determine next release version off a release branch
In more detail - this work is part of an effort to automate the release process by using a push to a release branch as a trigger to prepare, package and deploy react-native to npm from CircleCI
This function is later used in `prepare-package-for-release` script in D32556610 to bump the version
Reviewed By: cortinico, ShikaSD
Differential Revision: D32556609
fbshipit-source-id: 7d93ead0b34318a58ffeb876715fbd34d6041f4e
Summary:
The helper class for ScrollView should not need to detect the scroll direction. If it has to do so, that means the corresponding logic should live in the `ReactScrollView` or `ReactHorizontalScrollView` respectively.
This diff is to remove the scroll direction detection logic from the scroll view helper. Long term we should keep it this way as the shared code got moved to the helper class from the scroll view classes.
Changelog:
[Internal]
Reviewed By: javache
Differential Revision: D32514429
fbshipit-source-id: 2165f2eba90cc25d14834c39148fe8ce8805bea6
Summary:
This diff fixes two edge case (similar to a race condition) that caused unexpected behaviors.
**Problem one**
{F680816408}
The previous fling animation is not canceled when user starts to scroll or drag. This is causing both the animation and scroll are setting the scroll position. Depends on the animation path and scroll speed, there may be cases where the [velocity calculation](https://fburl.com/code/010lsu72) ends up getting reversed values. See P467905091 as an example where you can see `mXFlingVelocity` goes back and forth from positive to negative.
It's hard to see if the wrong values are in the middle, but if that happens in the end of user gesture, the velocity for the next fling would be wrong. It shows a "bounce back" effect, and can be triggered when user makes small quick joystick scrolls in one direction.
**Problem two**
{F680821494}
There is a gap between animator's `onAnimationEnd` lifecycle method [finished](https://fburl.com/code/6baq04ne) and the `Animator#isRunning` API to return false. This is causing issues for `getPostAnimationScrollX` where we [decide to return](https://fburl.com/code/hzzugvch) the animated final value or the scroll value. User may see the `-1` value got used for the next fling start value, and the whole scroll view goes back to the beginning of scroll view and starts to fling.
This happens when the previous fling animation finishes and the animated final value is set to -1, but at the same time the next fling starts before `isRunning` returns false for the previous animation.
**Solution**
The problems are fixed by
- Do not reset animated final value to -1 in `onAnimationEnd` method
- Add `mIsFinished` states and use it to track animation finish signal, instead of using `isRunning` API
- Update logic where we decide to return the correct value for the next animation starts point. We will return previous animated final value when the animation got canceled, and user is going towards that value from the current scroll value.
Changelog:
[Android][Fixed] - Fixed edge case for quick small scrolls causing unexpected scrolling behaviors.
Reviewed By: javache
Differential Revision: D32487846
fbshipit-source-id: f1b0647656e021390e3a05de5846251a4a2647ff
Summary:
If `value` is of type `float`, it will still fall through to the `react_native_assert` below, exploding. The `map` type is handled correctly.
Changelog: [Internal][Fixed] Fix fromRawValue(EdgeInsets) from single float case falling through to float array and exploding
Reviewed By: javache
Differential Revision: D32648848
fbshipit-source-id: e70cddd291a8f52d6ee3de5fef11b0bb7aee92cd
Summary:
The `__apply_Xcode_12_5_M1_post_install_workaround` script changes the `IPHONEOS_DEPLOYMENT_TARGET` to `11.0` for all pods. This causes problems if the pods were targetting `12.0` or higher. Many expo modules are targetting `12.0`.
I fixed this issue by checking the existing version and only bumping the target if it is lower than `11.0`.
See also: this discussion post by mikehardy https://github.com/reactwg/react-native-releases/discussions/1#discussioncomment-1619523
## Changelog
<!-- Help reviewers and the release process by writing your own changelog entry. For an example, see:
https://github.com/facebook/react-native/wiki/Changelog
-->
[iOS] [Fixed] - __apply_Xcode_12_5_M1_post_install_workaround causing pods targetting iOS 12 and above to fail
Pull Request resolved: https://github.com/facebook/react-native/pull/32633
Test Plan:
### Test (failing before this patch, passing after this patch)
1. pick an iOS Pod that has a minimum deployment target of iOS 12 or higher, I chose the Braintree package
2. `npx react-native init myrnapp`
3. Open `ios/Podfile` and add the pod as a dependency: `pod 'Braintree', '~> 5'` (and upgrade the Podfile target to 12 (`platform :ios, '12.0'`))
4. Compile the app.
Before applying this patch: ❌ Build fails because Braintree uses iOS 12 features and was downgraded to target 11.0
After applying this patch: ✅ Build succeeds
Reviewed By: fkgozali
Differential Revision: D32638171
Pulled By: philIip
fbshipit-source-id: 0487647583057f3cfefcf515820855c7d4b16d31
Summary:
The `add-version-label-issue` CI is failing if the version label is not found.
That is happening for previous versions of RN where the Labels are missing.
Changelog:
[Internal] [Changed] - Do not report a CI failure if add-version-label-issue is red
Reviewed By: sammy-SC
Differential Revision: D32723953
fbshipit-source-id: 878d2632b3a78311a01363b8f8a9181ae543a253
Summary:
As the title says. It's now safe to enable Fabric by default inside RNTester.
Changelog:
[Internal] [Changed] - Enable Fabric by default inside RN-Tester
Reviewed By: fkgozali
Differential Revision: D32723540
fbshipit-source-id: dab786694ba9ceb238570608713b5243a730aada
Summary:
Preallocation can take 10-20% of time when creating new nodes. (according to systrace measurements). At the moment, we are executing all preallocation calls in JS thread, potentially slowing down the progress there. Moving them to bg thread is a simple micro-optimization that ensures we return the node to JS as soon as possible.
Changelog: [Internal]
Reviewed By: sammy-SC
Differential Revision: D32677843
fbshipit-source-id: 3df47ae9aa365a8db720d71e2423c87659e9128c
Summary:
Removes extra .so files by merging built-in components into libfabricjni.so
These components shouldn't be referenced in outside modules, so merging them is trivial atm.
Changelog:
[Internal][Android] - Compile native components into static libraries
Reviewed By: cortinico
Differential Revision: D32677572
fbshipit-source-id: fc1a6c5a2832ee49e438c30856562f85677514ea
Summary:
Noticed we explicitly dropped the `allow_jni_merging` while not actually using it anywhere, and that we didn't have `fbandroid_allow_jni_merging` on some other libs.
Changelog: [Internal]
Reviewed By: RSNara
Differential Revision: D32355868
fbshipit-source-id: 6bd3fcc395e3dcacf4a8fc1033d471b2ffb0e8af
Summary:
Changelog: [internal]
Nothing was using `stop` parameter, let's get rid of it.
Reviewed By: philIip
Differential Revision: D32669018
fbshipit-source-id: dc2d52048a2f7dd3785dd959270087001c778962
Summary:
This task is supposed to really clean all the files that are generated
as part of a build and bring you to a clean state.
Changelog:
[Internal] [Changed] - Create a `cleanAll` Top Level Gradle task.
Reviewed By: ShikaSD
Differential Revision: D32649942
fbshipit-source-id: 20b72ad4e1c0ef046aaaba94e2331176dca49abf
Summary:
This diff makes the fling animator cuztomizable, so that the subclasses can have their own animation for fling behavior.
Before the diff, we rely on the `OverScroller` to `fling`, which has some issues with Spline interpolation being messed up due to the clamped fling distance (See more details in T105464095). This may not be a big issue for mobile when user touches screen, but in VR environment this shows up very clearly with joystick events. To fix that properly without affecting mobile behavior, I added a new interface `HasFlingAnimator` from the helper, and implemented with default fling animator in the OSS ScrollView.
We should consider adopt a suitable animator for mobile platform, as the non-smooth fling effect is also happening in mobile.
- Add interface `HasFlingAnimator` to `ReactScrollView` and `ReactHorizontalScrollView`
- Add default fling animator
- Depend on if the default animator is used, customize the flingAndSnap behavior
Changelog:
[Internal]
Reviewed By: JoshuaGross
Differential Revision: D32382806
fbshipit-source-id: 08f03350f6a9b9fc03414b4dcb9977b9f33603ba
Summary:
We have `LOCAL_SHARED_LIBRARIES` that are getting longer and are
making reviewing them on Diffs quite hard.
Having all the list of the dependency on a single line is suboptimal
and it makes hard to find duplicated entries.
I've updated the longest `LOCAL_SHARED_LIBRARIES` to be multilines and
I've sorted the entries here.
Changelog:
[Internal] [Changed] - LOCAL_SHARED_LIBRARIES
Reviewed By: ShikaSD
Differential Revision: D32695127
fbshipit-source-id: f5b381c501ddff083ef9f4baaca6c4c8c9523368
Summary:
Seems like the Gradle plugin was left behind when bumping AGP.
This bump is quite significant as AGP removed several dependencies from their
exported one so I had to reimport them again (ideally we should move to kotlinx-serialization).
I've also addressed a couple of Kotlin compiler warnings that were not related to
the AGP Api (those will be addressed at a later time).
Plus I've also fixed the target Java version to 11 as the compiler was complaining
that Java target was at 11 while Kotlin Jvm target was defaulted at 8
Changelog:
[Internal] [Changed] - Bump AGP to 7.x inside the react-native-gradle-plugin
Reviewed By: ShikaSD
Differential Revision: D32667745
fbshipit-source-id: 044930bf6cc49065eff4af1c9be79de76d5b368b
Summary:
Custom metro port not working without exporting variables from `.packager.env`
## Changelog
[General] [Fixed] - Fix support for custom port
Pull Request resolved: https://github.com/facebook/react-native/pull/32666
Test Plan: run `react-native run-android --port=8082` and `react-native run-ios --port=8082`
Reviewed By: yungsters
Differential Revision: D32694233
Pulled By: cortinico
fbshipit-source-id: 42e803d56b83608024b901d2a3024733ed7099b7
Summary:
This Diff fixes a crash happening as the user uses AppCompat 1.4.0
as a dependency in their App and uses a `TextInput` component.
The crash happens as `mFabricViewStateManager` is accessed during
the ctor of the superclass, and is not yet initialized.
Fixes#31572
Changelog:
[Android] [Fixed] - Fix crash on ReactEditText with AppCompat 1.4.0
Reviewed By: ShikaSD
Differential Revision: D32674975
fbshipit-source-id: efa413f5e33527a29fbcfa729e8b006ecb235978
Xin Chen
Luna Wei
Chiara Mooney
Liron Yahdav
zpd106
Yuan Dai
Sota Ogo
Samuel Susla
Tom Spencer
Marc Rousavy
David Ashinsky
Peter Carpenter
Janic Duplessis
Eric Rozell
Felipe Perez
Simon Farshid
Nicola Corti
Andrei Shikov
Pieter De Baets
Neil Dhar
enniel
