Summary:
Yarn key had expired on 2nd of February. The owner has renewed it, but seems like Circle CI still has the old one cached. (see https://github.com/yarnpkg/yarn/issues/6865#issuecomment-772047314)
Changelog: [Internal]
Reviewed By: fkgozali
Differential Revision: D26276386
fbshipit-source-id: 093c7de94445b6ff9beb5792b16564e5c3bd1234
Summary:
Changelog:
[Android][Changed] - Initial replacement of jcenter with mavenCentral.
Replaces jcenter with Maven Central in the build files of the repository.
Some dependencies are not resolvable from maven central yet, so for now they are included from jcenter, but limited to specific modules only.
I didn't touch the template for now.
Reviewed By: mdvacca
Differential Revision: D26260977
fbshipit-source-id: 2a65e1195b6367c026089184ad6471bf3573dc37
Summary:
Changelog:
[Internal][Yoga] - Added instance of checks in `YogaNodeJNIBase` class to prevent `ClassCastException`s. This was happening for some NT android tests - Mocked Yoga Node object was being passed in the `addChildAt` api
Stack Trace of exception
java.lang.ClassCastException: com.facebook.yoga.YogaNode$MockitoMock$1408896622 cannot be cast to com.facebook.yoga.YogaNodeJNIBase
at com.facebook.yoga.YogaNodeJNIBase.addChildAt(YogaNodeJNIBase.java:86)
at com.facebook.litho.DefaultInternalNode.addChildAt(DefaultInternalNode.java:220)
at com.facebook.litho.DefaultInternalNode.child(DefaultInternalNode.java:377)
at com.facebook.litho.DefaultInternalNode.child(DefaultInternalNode.java:360)
at com.facebook.litho.Column.resolve(Column.java:118)
at com.facebook.litho.Layout.create(Layout.java:172)
Reviewed By: Andrey-Mishanin
Differential Revision: D26114992
fbshipit-source-id: 774a689609e67f9244b81c6788b62cd61cd96d14
Summary:
LayoutAnimations: assert that tag >0 instead of != 0. It's possible that a corrupt tag value would be below zero which is not valid.
Changelog: [Internal]
Reviewed By: mdvacca
Differential Revision: D26271515
fbshipit-source-id: a62445ad29d60e5180e62ec4c6d5b08784655808
Summary:
We should always be able to reference a ComponentDescriptor. If there are any valid cases where the ComponentDescriptor doesn't exist (?) we need to document those and investigate more.
Changelog: [Internal]
Reviewed By: mdvacca
Differential Revision: D26271511
fbshipit-source-id: 659d160f82f1e78025d3dbe16efa0fa2d072d15f
Summary:
Make sure there's no corruption happening to these ShadowViews.
Changelog: [internal]
Reviewed By: mdvacca
Differential Revision: D26271509
fbshipit-source-id: 01bfa1bfce56e72b48304fe15b6f6426d3b5247e
Summary:
The new IntBufferMountItem queueing actually enforces a global ordering of mutation types: CREATEs, then INSERT, then REMOVE, then UPDATE, then DELETE.
See comments for more details. In general this ordering is fine, but if a DELETE animation is in progress and (due to view unflattening) the same view is recreated, the CREATE will be executed and then the in-process DELETE (since conflicting animations get flushed).
To mitigate this, in Binding we detect this and simply remove DELETE operations queued when we detect that we want to CREATE the node. `DELETE...CREATE` is valid but `CREATE...DELETE` in a single frame is not, so this is safe.
This does complicate and add more assumptions to Binding than I would like, but it should unblock us for now.
Changelog: [Internal]
Reviewed By: mdvacca
Differential Revision: D26271299
fbshipit-source-id: 9453fe17b8b541e484a047dc9637674dbdcc8e9a
Summary:
Publish a new react-native-codegen release with latest changes in order to support the 0.64 react-native release.
Changelog:
[Internal]
Reviewed By: fkgozali
Differential Revision: D26253312
fbshipit-source-id: 766cfa3e088b3715bc9bc815523c8d04fae409bf
Summary:
Currently, there is a bug with the way that fabric translates float values for color components to hex/int value (which android uses). There are two main problems with the current setup. The first is that we are using 256 as our ratio instead of 255 which is the maximum value for any color component in hex. The second is that we cast the components to ints instead of rounding which truncates the values. We uncovered this with UIQR because our off-by-one color values were incorrectly off a little bit extra for fabric surfaces.
Changelog: [Internal] updates the float-to-int value conversion for color components to be fully accurate
Reviewed By: shergin
Differential Revision: D26161396
fbshipit-source-id: 884e27ffa01b116f9307c00298bb8d888f9f6dd7
Summary:
FabricUIManager does not support it, the declaration is not correct.
Changelog: [Internal] Fabric-specific internal change.
Created from Diffusion's 'Open in Editor' feature.
Differential Revision: D26241483
fbshipit-source-id: 8a894dc847bce9c196d8ac2e1601853e4fe03e1d
Summary:
This is a Fabric-compliant implementation of `JSResponder` feature. To make it work e2e we also need to update FabricRenderer in React repository. But before we can do this, we need to ship the native changes.
Changelog: [Internal] Fabric-specific internal change.
Reviewed By: mdvacca
Differential Revision: D24630027
fbshipit-source-id: 70c30e1250b554d83862956b536714704093072f
Summary:
This is an example from RNTester. I added a button in the ScrollView which can eat taps. This allows us to test the `keyboardShouldPersistTap` prop values for `ScrollView`, since there was no way to differentiate between "handled" and "always".
Expected behavior for each value of `keyboardShouldPersistTaps`:
* `"handled"` - The keyboard should dismiss if the ScrollView is tapped, but pressable components like buttons can eat taps (and not dismiss the keyboard)
* `"never"` - The keyboard should always dismiss when anything in the ScrollView is tapped. Buttons won't be pressed - the keyboard will dismiss instead.
* `"always"` - The keyboard should never dismiss. Buttons and other content in the ScrollView can be tapped.
Reviewed By: lunaleaps, nadiia
Differential Revision: D26228852
fbshipit-source-id: 1df78f0be10e2f2cc81e95548a6a6f7cab07894f
Summary:
Simplifies `AppState` by removing redundant methods and changing `addEventListener` to return an `EventSubscription`.
Changelog:
[General][Changed] - `AppState.addEventListener` now returns an `EventSubscription` object.
[General][Removed] - Removed `AppState.removeEventListener`. Instead, use the `remove()` method on the object returned by `AppState.addEventListener`.
[General][Removed] - `AppState` no longer inherits from `NativeEventEmitter`, so it no longer implements `addListener`, `removeAllListeners`, and `removeSubscription`.
Reviewed By: wtfil
Differential Revision: D26161343
fbshipit-source-id: b3cff76bf0f8f7d79cd954fdef551d0654c682ca
Summary:
Changelog: [Internal]
This removes a couple dependencies on bridge in dev menu native module. So far reloading works, will get inspector and other options working in follow ups.
Reviewed By: sammy-SC
Differential Revision: D26240391
fbshipit-source-id: 7c9d585f4efa6cc9db995ef8a33865831bc8d526
Summary:
Live reload was removed as a user facing feature in 2019 in favour of fast refresh. The native code was left in case "automation" relied on it. I'm quite sure no automation is using this dev feature. Remember, the dev feature made it so that hitting save in a text editor auto reloaded in metro.
Cleanup the native implementation.
Changelog: [Internal]
Reviewed By: sammy-SC
Differential Revision: D26239628
fbshipit-source-id: 7f61c7204727bb2d739600a459f69c72842265c5
Summary:
IIUC `isDebuggingRemotely` will never be set when `isDeviceDebuggingAvailable` is true. This was just a failsafe so that noone was trapped in an intermediate state.
Changelog: [Internal]
Reviewed By: sammy-SC
Differential Revision: D26239570
fbshipit-source-id: 8db6b022a0ea581216a1fa97b9d41f5ab6160562
Summary:
Changelog: [internal]
Provide a default plugin function that provides core `RCTModalHostViewComponentView`. This makes for easier setup for standalone app migration.
Reviewed By: JoshuaGross, shergin
Differential Revision: D26150091
fbshipit-source-id: d39723b99c590d1a33fb7b628f809aa12b2f7589
Summary:
Removes the checks in `RCTDeviceEventEmitter` that were initially added to migrate call sites to other modules when `NativeEventEmitter` was introduced.
This set of checks is problematic for a few reasons:
1. Does not cover many other events that have since been introduced, so the small list covered is arbitrary.
2. Prevents composition of `RCTDeviceEventEmitter` in the implementation of those modules.
3. Code bloat.
Changelog:
[General][Removed] - `RCTDeviceEventEmitter` no longer throws for `SttatusBar`, `Keyboard`, and `AppState` events. However, you are still recommended to use the more appropriate modules for listening to these events.
Reviewed By: lunaleaps
Differential Revision: D26163602
fbshipit-source-id: 316287bfdf2947fe85d022a3f83a205e89c432ba
Summary:
## Problem:
NativeModules can only be initialized after we mark them initializable on the NativeModules thread. This work is scheduled in ReactInstanceManager.setupReactContext(), after we schedule the execution of the JS bundle on the JavaScript thread in ReactInstanceManager.createReactContext():
https://www.internalfb.com/intern/diffusion/FBS/browse/master/xplat/js/react-native-github/ReactAndroid/src/main/java/com/facebook/react/ReactInstanceManager.java?commit=9b58f58b8eb12281567b8d24d6d000e868e61a1f&lines=1256-1257%2C1331%2C1334-1335%2C1333-1334
So, if the timings don't work out, the JavaScript thread could start executing the JS bundle before the NativeModule thread makes all NativeModules initializable. In this case, those NativeModule will just appear to be null in C++/JavaScript.
## Fix
This diff makes all NativeModules initializable immediately after their ModuleHolder is created. ModuleHolder.markInitializable() simply initializes initializes modules that were eagerly created.
Changelog: [Android][Fixed] - Make NativeModules immediately initializable
Reviewed By: JoshuaGross
Differential Revision: D26233372
fbshipit-source-id: a9223ff093da5b80781169be88e6ec9516c7a29b
Summary:
The windows CI tests are flaky due to network failures trying to install the android SDK. This modifies the CI to retry a couple of times if the first install fails.
CircleCI will timeout the command if there is no output to the console for 10mins. I've added a timeout to the choco command for 9mins, which should cause it to timeout before the circleci one (assuming that it spends a non-zero time with output before hanging on the download) On successful downloads runs the download seems to be substantially quicker than the 10min timeout.
I added --force and --forcedependencies on the retries, otherwise if the jdk package fails, but the androidsdk package installs, the retry will still not install the jdk package.
This doesn't fully fix the issue, since the CI is still hitting the network, which could be flaky all 3 attempts. But hopefully it'll improve the success rate somewhat. -- Really the Windows CI job should have some kind of docker image or something that already has all the prereqs installed.
## Changelog
[Internal] [Fixed] - Windows CI Android SDK install retry
Pull Request resolved: https://github.com/facebook/react-native/pull/30742
Test Plan: We'll see if the windows CI becomes more reliable
Reviewed By: fkgozali
Differential Revision: D26231499
Pulled By: hramos
fbshipit-source-id: f4a7f2132b3317f4763ddadcd909ef1c2a43dc2f
Summary:
This diff refactor strong references on TM infra to use weak references and avoid leaking objects
changelog: [internal] internal
Reviewed By: RSNara
Differential Revision: D26229814
fbshipit-source-id: d4a327dba227378d23764433d5917eb4378a4453
Summary:
In bridegless mode, `setBridge` is never called on native modules. This diff refactors out some setup logic into the module's init method.
Changelog: [Internal]
Reviewed By: RSNara
Differential Revision: D26211620
fbshipit-source-id: 7a9910198852dea5ac679762aa93dfea6fc47d70
Summary:
When NativeUIManager.getConstantsForViewManager throws an exception from native, we just swallow that exception and set null as the view config. Instead, we should log that an error occurred. This way, we can tell if the NativeModule sync method call returned undefined, or if the NativeModule sync method call thew an exception.
I'll take a look and see if we can add this logging higher up in the native stack, so that we can actually capture the stack information from this NativeModule sync method call failure.
Created from Diffusion's 'Open in Editor' feature.
Changelog: [Internal]
Reviewed By: mdvacca
Differential Revision: D26227379
fbshipit-source-id: 02b516d247f82f873f059005e6a2cc6f2a64fdb5
Summary:
JSI had its compiler options like `-Wno-non-virtual-dtor` exported to targets
that depended on the JSI library.
These flags should be private to the compiler invocation on the `jsi.cpp` translation
unit and don't need to be shared with any libraries pulling in `jsi.h`.
Fixes#448
Changelog:
[Internal] - Don't export C++ warnings from JSI CMake files
Reviewed By: Huxpro
Differential Revision: D26212636
fbshipit-source-id: bd95580f89694189975a49660c439e4e5f7c5afc
Summary:
How does an application register a TurboModuleManagerDelegate with ReactInstanceManager?
1. Call ReactInstanceManagerBuilder.setReactPackageTurboModuleManagerDelegateBuilder(ReactPackageTurboModuleManagerDelegate.Builder)
2. Override ReactNativeHost.getReactPackageTurboModuleManagerDelegateBuilder()
Changelog: [Android][Added] - Introduce API to allow applications to register TurboModuleManagerDelegates with ReactInstanceManager
Reviewed By: mdvacca
Differential Revision: D26193055
fbshipit-source-id: bf82e63e6ab1c0c8f12bada92ac6852c992ec9cb
Summary:
We'll pass these ReactPackageTurboModuleManagerDelegate.Builders to ReactInstanceManager. ReactInstanceManager will then use these builders to initialize the TurboModule system. To allow TurboModules to work with RNTester, we have to switch RNTester over to using these Builders.
Changelog: [Internal]
Reviewed By: mdvacca
Differential Revision: D26193054
fbshipit-source-id: d98b857226be3ebaa22b1b734a4e76c0f4be6608
Summary:
Previously, owner(TurboModuleManager) used to depend on owner(ReactInstanceManager). Now, owner(ReactInstanceManager) depends on owner(TurboModuleManager).
**Rationale:** This allows ReactInstanceManager to create TurboModuleManager.
## Changes
We moved ReactPackageTurboModuleManagerDelegate to com.facebook.react.
Changelog: [Internal]
Reviewed By: mdvacca
Differential Revision: D26163661
fbshipit-source-id: 3ebff16ef2aa77e20bb55500ed44c9214acb91dd
Summary:
Deletes `Linking.removeEventListener` in favor of returning the `EventSubscription` returned by `NativeEventEmitter.addListener`.
Changelog:
[General][Added] - `Linking.addEventListener` now returns a subscription with a `remove()` method.
[General][Removed] - Removed `Linking.removeEventListener`. Instead, call `remove()` on the subscription returned by `Linking.addEventListener`.
Reviewed By: rubennorte, wtfil
Differential Revision: D26155896
fbshipit-source-id: 1176ec6eae1c0fff2d68acf5411a18da530b212a
Summary:
Currently, Codegen bash wrapper (`generate-specs.sh`) for Xcode invokes JS-based Codegen tooling via `yarn --silent node <...>`. This breaks both:
- when Yarn is not installed (if NPM is used), for obvious reasons
- when Yarn v2 ("Berry") is active
This PR changes the way `generate-specs.sh` locates `node` executable to the following algorithm:
- use the path provided in the `NODE_BINARY` env var
- if `NODE_BINARY` env var is not defined, find `node` with `command -v node`
## Changelog
[iOS] [Fixed] - Fix Codegen silently failing when Yarn is not installed, or when Yarn v2 is active.
Pull Request resolved: https://github.com/facebook/react-native/pull/30781
Test Plan:
### Case 1 (no Yarn installed)
1. Ensure `yarn` is not present in PATH
2. Run Xcode build
3. Check that Codegen artifacts are produced
### Case 2 (Yarn v2 is used)
1. Ensure `yarn` is running in the v2 ("Berry") mode
2. Run Xcode build
3. Check that Codegen artifacts are produced
Reviewed By: fkgozali
Differential Revision: D26187081
Pulled By: hramos
fbshipit-source-id: 77d3089f523b8c976d8223b77ff9553cb6cf68a5
Summary:
Tiny PR that removes an old orphaned `.nvmrc` file (added in 2018 and not touched since then).
Most likely the project does not work with Node.js 8.11.3 anymore. There is (no longer) a need to attempt to strictly enforce a specific version client-side. `package.json` _already_ declares
```
"node": ">=12"
```
in the `engines` section, so we should be good.
## Changelog
[General] [Removed] - Delete .nvmrc
Pull Request resolved: https://github.com/facebook/react-native/pull/30765
Test Plan: Nothing to test...
Reviewed By: hramos
Differential Revision: D26190114
Pulled By: TheSavior
fbshipit-source-id: 515807da8a9e28fd28c12e8b5416ed125436341c
Summary:
ES Modules implicitly enable strict mode. Adding the "use strict" directive is, therefore, not required.
This diff removes all "use strict" directives from ES modules.
Changelog:
[Internal]
Reviewed By: motiz88
Differential Revision: D26172715
fbshipit-source-id: 57957bcbb672c4c3e62b1db633cf425c1c9d6430
Summary:
Changes the to-be-legacy `_EventSubscription` and `_EmitterSubscription` classes to implement the event-agnostic `EventSubscription` interface.
This interface is valuable because it abstracts away the event definitions.
Changelog:
[Internal]
Reviewed By: lunaleaps
Differential Revision: D26155911
fbshipit-source-id: f94280976d4f9219f4780ac8fae0d5fbc1865386
Summary:
Simplifies the implementation of `DevSettings`. The fact that it uses `NativeEventEmitter` is an implementation detail and does not need to be exposed via inheritance.
This also enables more code to be removed from production builds (because the development implementation is now statically enclosed by `__DEV__`).
Changelog:
[General][Changed] - `DevSettings` no longer inherits from `NativeEventEmitter`
Reviewed By: RSNara
Differential Revision: D26140148
fbshipit-source-id: 99fa9d0c6ce8e365f89936aa12a4720f7a04b984
Summary:
Removes the `options` argument in the `NativeEventEmitter` constructor. It was only being used for an experimental flag that is no longer necessary.
Changelog:
[General][Removed] - Removed second optional argument of `NativeEventEmitter` constructor
Reviewed By: RSNara
Differential Revision: D26138239
fbshipit-source-id: 0481ce44f0464668e3a6e7ffaf079d17c87afd42
Summary:
This addesses a few issues I noticed while migrating my app to the new build-time codegen on iOS.
1. I noticed random failures because of codegen on iOS. This is mostly due to the fact the codegen output files are not specified in the xcode script. The only reason it works relatively fine currently is because the codegen output is inside the input files directory. This has the side effect of causing files to be regenerated every build, then causes all core modules to be recompiled which adds up a significant amount of time to rebuilds. To fix this I added the generated files to the script phase output and moved the FBReactNativeSpec dir outside of the codegen source (Libraries). I moved it to the React directory as this seemed to make sense and is where a lot of iOS files are as well as the core modules. Note this might require internal changes. This removes the circular dependency between our build phase input and output so consecutive builds can be cached properly.
2. Add `set -o pipefail` to the xcode script, this helped propagate errors properly to xcode because of the `| tee` pipe so it fails at the script phase and not later with a header not found error. Also add `2>&1` to pipe stderr to stdout so errors are also captured in the log file.
3. Add the `-l` flag to the bash invocation to help finding the yarn binary. With my setup yarn is added to the system PATH in my user .profile. Adding this file will cause bash to source the user environment which xcode scripts does not by default. I think this will help with most setups.
4. If yarn is not found the `command -v yarn` would make the script exit without any output because of the -e flag. I made a change to ignore the return code and check later if YARN_BINARY is set and have an explicit error message if not.
## Changelog
[iOS] [Fixed] - Make codegen more reliable on iOS
Pull Request resolved: https://github.com/facebook/react-native/pull/30792
Test Plan:
Tested various project states to make sure the build always succeeds in RN tester:
- Simulate fresh clone, remove all ignored files, install pods, build
- Build, delete FBReactNativeSpec generated files, build again
- Build, build again, make sure FBReactNativeSpec is cached and not rebuilt
- Make the script fail and check that xcode shows the script error logs properly
Note: Did not test fabric
Reviewed By: fkgozali
Differential Revision: D26104213
Pulled By: hramos
fbshipit-source-id: e18d9a0b9ada7c0c2e608d29ffe88087f04605b4
Summary:
See comments. On iOS in Fabric, LayoutAnimations is only conditionally enabled; whereas on Android it's always enabled. That means for code on iOS that relies on the onComplete callback, there might be bugs.
Ensure the callback is always called on iOS by racing a timer with the animation completion.
This will be deleted before Fabric "ships" fully.
Impact is minimal since this change is scoped to only run on iOS and under Fabric.
Changelog: [Internal]
Reviewed By: sammy-SC
Differential Revision: D26166237
fbshipit-source-id: 9a07a402845c379e1511f199eef3d3e249e1eb06
Summary:
Changes:
* RCTMountingManager got two dedicated methods that designate some external view as a Surface viewport;
* The wiring between views now happens right in `start` and `stop` methods (because there is no reason to do it later); now it's symmetrical and straightforward;
* A couple of Surface legacy statuses were removed; they are not used and do not convey useful information anyway.
Changelog: [Internal] Fabric-specific internal change.
Reviewed By: sammy-SC
Differential Revision: D26095434
fbshipit-source-id: 37828e093a40a1ad22ad78f110c18a8f27781ba7
Summary:
Refactors `RCTNeworking.ios.js` so that event registration does not get passed along to `NativeNetworkingIOS`. Instead, we go straight to `RCTDeviceEventEmitter` (which is what `NativeEventEmitter` ultimately does when `nativeModule` is not supplied).
This optimization reduces overhead of making network requests, and it is made possible because `NativeNetworkingIOS` does not actually do any meaningful work when `startObserving` is invoked.
Changelog:
[iOS][Removed] - Removed event methods except `addListener` from `Networking`
Reviewed By: lunaleaps
Differential Revision: D26137965
fbshipit-source-id: b6e0288689459ddb8ecf8d34dce6250d3b0ecb59
Andrei Shikov
Sidharth Guglani
Joshua Gross
Héctor Ramos
Tim Yung
David Vacca
Sara Valderrama
Valentin Shergin
Kacie Bawiec
Peter Argany
Samuel Susla
Ramanpreet Nara
Andrew Coates
Lulu Wu
Riley Dulin
Ivan Moskalev
Frieder Bluemle
Micha Reiser
Janic Duplessis
