Summary: Developer tools have a lot of dependencies on the bridge, so for now I'm just disabling them in bridgeless mode entirely by guarding everything in `setUpDeveloperTools` with the `RN$Bridgeless` flag. Also consolidating some of the stuff Dan added to `InitializeCore` for hot reloading in here.
Reviewed By: zackargyle
Differential Revision: D15797924
fbshipit-source-id: 360ea81a2844e49f7281eed259fc16a541148ac2
Summary:
This diff reimplements the prop parsing infrastructure in a part where it interacts with RawProps value.
Local synthetic tests show that the new way is 3x faster but the actual production result is quite unpredictable. MobileLab tests show some improvements about 10-20 ms on iPhone 6.
In short, the new way is faster because it inverts the lookup order and heavily relies on actual data types (and their properties) that we use. The old approach required about 130 hash-map lookups (where the key is `std::string`) to parse a single *Props object.
The new approach prepares concrete-props-specific tables with indexes of coming values ahead of time, iterates over raw data and puts it into those tables, and then performs a lookup in a very efficient manner.
Reviewed By: JoshuaGross
Differential Revision: D15752968
fbshipit-source-id: 847106e652eb7fc7ef7b99884a6f819ea3b9fd06
Summary: Even though it makes no sense to construct RootProps from RawProps, we need to support that (even if it's no-op) to be able to call this once from generic `ConcreteComponentDescriptor`. We will need it in the coming diff.
Reviewed By: mdvacca
Differential Revision: D15752970
fbshipit-source-id: b75a4023c5d0425a8dbe0f104a36a0f265eb6084
Summary:
In theory, those annotations can help the compiler to emit more optimized code and/or code that suggest CPU the proper way to utilize the instruction pipeline after the coming branch.
TBH, it's not clear how exactly beneficial those annotations are (esp. on target architectures) but they certainly don't hurt (in all cases here the favorite code branch is obvious).
Reviewed By: mdvacca
Differential Revision: D15752969
fbshipit-source-id: 8adf25a48107ffde828f735fb1386b30dbe63ede
Summary:
Previously we used `std::string` as a type behind all prop names (and name fragments). Even if `std::string` converted from `char const *` should be heavily optimized by STL and compiler, we still concatenate and copy those strings a lot. Switching to `char const *` allows avoiding tons of copying and inefficient equality checks.
Besides that, the future, more sophisticated optimization will rely on this.
Reviewed By: mdvacca
Differential Revision: D15511493
fbshipit-source-id: 9f509d18f0c737f7f77d4fea192d2ed1872e3731
Summary: This is a small perf tweak that alone does not give much improvement but the coming diffs will rely on the possibility to construct empty RawProps object.
Reviewed By: mdvacca
Differential Revision: D15511494
fbshipit-source-id: 39dec9336e6b5cf6ad33b1f3a06ca1c096ece628
Summary: Setting value directly to ivar makes impossible to subclasses to react of this; this is a bug.
Reviewed By: mdvacca
Differential Revision: D15776691
fbshipit-source-id: 10f9be975ee673b2db46c24ba41b18f6d6ddf0a3
Summary:
In some cases, the view class is the only that retains stored `props_` variable. At the same time the `[super updateProps:props oldProps:oldProps];` actually resets the pointer with a pointer to new props value which sometimes causes the deallocation of the old value. All that is okay unless the implementation of `updateProps:oldProps:` in superclasses stores a raw reference to an old value in the very beginning of the method (for convenience and perf reasons).
So, to prevent preliminary deallocation of the old value pointed by `_props` we moved all `[super updateProps:props oldProps:oldProps];` calls to the end of overloaded methods.
Reviewed By: mdvacca
Differential Revision: D15770068
fbshipit-source-id: af36b3e70560ab00846cd26b0963bbc059e977bc
Summary: It's a antient left over which was copy pasted a dozen of times. All `RCTViewComponentView` subclasses have own copy of the old props `_props`, so it's useless to check for `oldProps`.
Reviewed By: mdvacca
Differential Revision: D15770067
fbshipit-source-id: 39f4d71ccdcf0c9a5b911b17a3b922dbe6dd9a8e
Summary:
Address Sanitizer told me that I have UB in `RCTUIFontWeightFromFloat` and it was right. After a short investigation, I find out that the original assumption that `UIFontWeight` values are practically numbers from 100 to 900 was incorrect. In my simulator, it's something like from `-1` to `+1` (irregularly!). So, the whole subsystem worked only by accident.
This diff fixes that; now we never assume which actual values `UIFontWeight` constants have.
I will publish code style fixes as a separate diff (otherwise it will be really hard to review).
Reviewed By: JoshuaGross
Differential Revision: D15756620
fbshipit-source-id: d7f888e85815d863487c6b68a960e39fd473e095
Summary:
This PR only fixes a one small orthographic error
Pull Request resolved: https://github.com/facebook/react-native/pull/25245
Differential Revision: D15796478
Pulled By: hramos
fbshipit-source-id: ccb811b43b0d2efc5d97ba335b60531a0fcbda10
Summary:
It turns out that it's expected in certain cases for `UIManager.getViewManagerConfig` to return null: https://fburl.com/4h4pqtd7
Instead of throwing when you try to call that function, let's log something and return null.
Reviewed By: fkgozali
Differential Revision: D15791367
fbshipit-source-id: 71cf14071d877070b4f8b2d72eaa2f10beac38db
Summary: For bridgeless RN we're not going to use the JSTimers module + Timing native module for timers (e.g. setTimeout, setImmediate, etc.). Instead we're going to install global functions from cpp for each of these, so we can just skip setUpTimers entirely in this case.
Reviewed By: fkgozali
Differential Revision: D15790638
fbshipit-source-id: 1626fe90a27cb8d385cbb700ad932969f708f0cb
Summary:
Not totally sure if this is the best way to handle this. In Venice if a native module is missing I try to log the name of the module, but I noticed that the error I was getting was getting this:
{F161460962}
Presumably this is because importing from NativeModules looks for `__esModule`, but NativeModules uses `module.export`. So it's trying to access that property on my cpp proxy object, which doesn't exist...? Changing TurboModuleProxy to use `require` seems to fix the problem.
Reviewed By: fkgozali
Differential Revision: D15787508
fbshipit-source-id: 4b9df4e3c179117999fe6de6363edbef427a8263
Summary:
Replacing UIManager.js with a shim that redirects to either PaperUIManager (containing old impl) or DummyUIManager, if `global.RN$Bridgeless` is set. The UIManager native module doesn't exist in bridgeless mode, which means requiring UIManager.js currently fatals. This is a bit hacky, but it's a lot easier than implementing a dummy native module to make it happy.
I did have to stub out all the properties in UIManagerJSInterface to appease flow, though...
Reviewed By: yungsters
Differential Revision: D15775582
fbshipit-source-id: 8e2628f75b2242971895583696122760acdad7af
Summary:
This diff adds codegen support for flow types such as:
```
type ModuleProps = $ReadOnly<{|
size: $ReadOnlyArray<('small', 'large')>
|}>
```
These array enums are codegen'd as bitmaps in c++
Reviewed By: sammy-SC
Differential Revision: D15766763
fbshipit-source-id: 8c55303fb3a0ab151eae2b441119ab078e2c5d3d
Summary: This is a temporary backout of D14817454, to verify if this is related T45503571
Reviewed By: JoshuaGross
Differential Revision: D15780018
fbshipit-source-id: 455034ce7b7096101db93a8604b77e1233db1137
Summary:
Using Animated components like View and Image do not get created with __skipSetNativeProps_FOR_TESTS_ONLY = true since they get created before the jest mock can be applied to createAnimatedComponent. For these components mock getters to create versions that properly skip set native props.
Jest tests that render these components get warnings the setNativeProps gets called even though using a testRenderer this should not happen.
## Changelog
[Javascript] [Fixed] - Define Animated Components for react-native/jest/setup that properly skip setNativeProps
Pull Request resolved: https://github.com/facebook/react-native/pull/25108
Differential Revision: D15779434
Pulled By: cpojer
fbshipit-source-id: f39e21ed8e71c2c155297c791d3bf573909896d6
Summary: Fishhook was used to try to hide the log messages from RCTReconnectingWebSocket but that doesn't really work anymore. Deleting it now to unblock people trying to build for iOS 13.
Reviewed By: cpojer
Differential Revision: D15779390
fbshipit-source-id: ef18575d5d92ac374e189b1267dee3a9befc3551
Summary:
Modify the wrong word in the comment of NativeViewHierarchyManager.java.
Fix the explanation of the code to avoid misunderstanding.
## Changelog
[Android] [Fixed] - Fix the explanation of the code to avoid misunderstanding.
Pull Request resolved: https://github.com/facebook/react-native/pull/25235
Differential Revision: D15779249
Pulled By: cpojer
fbshipit-source-id: d4e7baa9ea9be5551feed8f643fe9774b3226bd8
Summary:
Passing Measure callback data - width, widthMeasureMode, height, heightMeasureMode, measuredWidth and measuredHeight along with NodeMeasure event
This data is then propagated to java layer in this diff
Reviewed By: davidaurelio
Differential Revision: D15697523
fbshipit-source-id: 615463da237175ff88abef3f6528b55333ccd915
Summary: This makes it so that you don't have to specify the package to the bundle. When you don't specify it, it will use the default bundle that is chosen when the app starts.
Reviewed By: PeteTheHeat
Differential Revision: D15742001
fbshipit-source-id: cd5906a66d04eea308dbc0516bc1fec60caed6bb
Summary:
`fbsource//xplat/js/react-native-github/ReactAndroid/src/main/java/com/facebook/react/turbomodule/core/jni:jni` is a shared library that depends on `fbjni` (also another shared library). In dev and prod, I think `turbomodule/core/jni:jni` was pulling in `fbjni` from its `react_native_target("jni/react/jni:jni")` dependency even though `FBJNI_TARGET` wasn't an exported_dep of `react_native_target("jni/react/jni:jni")`. In `android_automation_cameracore_release_build`, however, this wasn't happening, which led to that build failing to compile.
In this diff, I'm explicitly making `FBJNI_TARGET` a dependency of `fbsource//xplat/js/react-native-github/ReactAndroid/src/main/java/com/facebook/react/turbomodule/core/jni:jni`. This unbreaks `android_automation_cameracore_release_build`.
Reviewed By: fkgozali
Differential Revision: D15758160
fbshipit-source-id: db20c127a888b0cf68aea592de6a2e380c455c1d
Summary:
The necessity of this feature was removed in 2017. To intercept React Native gesture recognizer, implement UIGestureRecognizer delegate for conflicting gestures.
Here is the quote from the internal note:
> Previously we had lots of super weird bugs where React Native would inaccurately recognize touch gestures which were meant to be addressed by the native environment. Usually, these bugs occurred as unintentional taps happening right after swipe gestures. In all of these cases, we had to manually call method `cancelTouches` as part of an external gesture recognition process which prevented touch delivery to React Native. Furthermore, we had to delay touch delivery to React Native to wait for these cancellations. That code always looked like a hack (in the bad meaning of this word), like in some random place something dispatch event to another random place where something finally calls `cancelTouches`, yak. It was super annoying because it required adding this hack to all existing apps and screens, and because sometimes it was even too late to cancel touches.
> We fixed that. Instead of delaying touch delivery and waiting for calls to `cancelTouches`, we set up the React Native gesture recognizer in such way that it always agrees to fail in favor of any native gestures (from non-RN-based and served views which are placed higher in a hierarchy). React Native will now cancel all active touches itself so that we no longer need to call `cancelTouches` manually. We already removed all these calls and supported code from Facebook and Instagram.
See also: https://github.com/facebook/react-native/pull/25193
Reviewed By: PeteTheHeat
Differential Revision: D15734129
fbshipit-source-id: 289f77a437cb40199c591153b5801d24d0c10d1e
Summary:
This diff adds support for:
```
propName: WithDefault<string, null>,
```
It will throw if null is used for any other type like boolean
Reviewed By: TheSavior, cpojer
Differential Revision: D15748556
fbshipit-source-id: 925457ca1739bfad08e4776ecb47c0beb3acacf5
Summary: This diff turns on codegen for Slider and Switch
Reviewed By: TheSavior
Differential Revision: D15738544
fbshipit-source-id: a0dfb5b05fd62f28fc3865855986e49598dd5e19
Summary:
This diff adds a testing screen dev route to the facebook app for testing generated view configs
It's not pretty (i have 0 tetra experiance) but it gets the job done
There are three cases handled:
- No generated config �
- Invalid generated config (useful for dev) �
- Valid generated config �
On the description page we:
- Redbox it it's invalid (this could be used to redbox test all host components)
- Show diffs of the view config properties
- List all of the generated config properties
- List all of the native config properties
Using this tool, it's easy to see what the current config on native is, add correct flow types for the generated config, and validate the generated config
Coming later: adding all of the native configs to the list (will probably need filtering)
Reviewed By: cpojer
Differential Revision: D15683033
fbshipit-source-id: 5a566a56bef4f3f0bac3ea581c2e6acb2b9984e3
Summary: D15753278 brokes the build on the armv7 arch. Just backing out this diff and the build works again.
Reviewed By: rzito
Differential Revision: D15758272
fbshipit-source-id: 4e3d3f5322346d31d6160b66b8fef15963baec83
Summary:
This module is being removed from React Native via Lean Core. This diff moves all the related JS files to FB internal.
Note: I removed two references to previously removed modules from some files in this diff. I hope you don't mind.
Reviewed By: TheSavior
Differential Revision: D15714919
fbshipit-source-id: 88ea406396b31f5c255e06d9c92b67127c81db4a
Summary: This module is being removed from RN as part of the Lean Core effort.
Reviewed By: TheSavior
Differential Revision: D15714507
fbshipit-source-id: bb5dc2025a25ad450d6971e5948e7a2e678a9a25
Summary: An attempt to integrate the module flow types with internal codegen infra. Nothing of interest here, other than minor tweak on a spec (we don't support tupples...).
Reviewed By: mdvacca
Differential Revision: D15753278
fbshipit-source-id: b91d564fdbe8f72b90bea725779a9684993472b5
Summary: Some native modules methods expects number-based args like `NSDate`. For backward compatibility, the incoming numbers should be converted using RCTConvert, just like object args.
Reviewed By: mdvacca
Differential Revision: D15748968
fbshipit-source-id: 4db2cb0c41eda1bbe8cde7b0365d9c3d675f5fb5
Summary: Some native modules defined `synthesize methodQueue` which will ask the bridge to create a new serial queue for them. TM system needs to preserve this behavior for backward compatibility. In the future though, this magic shall be removed.
Reviewed By: mdvacca
Differential Revision: D15748198
fbshipit-source-id: 66a4b60a2769ac967a8d3bb00c4c635a68daebbc
Summary: `verifyComponentAttributeEquivalence` checks the new JS view configs against what we get from native at runtime (in dev only). This breaks in bridgeless mode because there is no paper UIManager to ask for the viewconfigs from; this diff uses the flag from D15721940 to skip this check.
Reviewed By: fkgozali
Differential Revision: D15722127
fbshipit-source-id: d9227107e0ff7814c34beaae6461bb8232699c94
Summary:
In bridgeless mode we don't want to set up the batched bridge, which is set up as part of InitializeCore. Instead of deleting InitializeCore completely, let's just skip this step if we're in bridgeless mode, which we'll detect using a global variable set on the runtime from cpp (`RN$Bridgeless`).
This way you still get an error if the bridge is somehow not set up properly when you're not in bridgeless mode (it won't fail silently).
Reviewed By: fkgozali
Differential Revision: D15721940
fbshipit-source-id: 73896e25874dd000f37d1abc9cf6be549ab3434f
Summary:
D14014668 introduced support for nesting views within Text on Android. Part of the implementation involved accessing the UIManagerModule from ReactTextView through context. This doesn't work in bridgeless RN because we have no UIManagerModule, and the ReactContext has no Catalyst instance. Trying to access the Catalyst instance from ReactContext throws an exception if it doesn't exist, so i'm just adding a simple check here to make sure the instance exists before proceeding.
This means that this feature won't work in bridgeless mode, but that's ok for now - eventually we want to change the way this works so that it doesn't rely on accessing views in Java, which is potentially unsafe (there's nothing to stop you from mutating the views, and cpp/js would never know about it).
Reviewed By: mdvacca
Differential Revision: D15703100
fbshipit-source-id: 0448d55b8345fc707a25210a505cb6ac520c708a
Summary: This diff removes the `React.js` forwarding module from RN open source which was only used for haste requires. I moved this file to FB internal and manually changed requires from `React` to `react`. However, I can't fully remove this forwarding module because we have files shared from www that expect to require React via the capitalized named.
Reviewed By: yungsters
Differential Revision: D15738887
fbshipit-source-id: b5b6c0be258582cfad92c13d174e5490c75152d9
Summary:
Reverse the order in which we look for modules in TurboModuleRegistry to check TurboModules first, and then fall back to legacy native modules. The main motivation for this is Venice, since requiring NativeModules.js fatals because there's no batched bridge. But we'll probably want to do this eventually anyway.
I ran a mobilelab for Marketplace home and am not seeing any significant difference in TTI.
Reviewed By: fkgozali
Differential Revision: D15703655
fbshipit-source-id: d65a4d7e09077474c30fb3938e38aee63bfa4eca
Summary:
These changes were originally landed in D15327683. Unfortunately, I had to back these out because they revelaled problems with our TurboModules infra. Those problems have since been fixed, so I think it's safe for us to re-land these changes.
Please refer to D15327683 for a description.
Differential Revision: D15739355
fbshipit-source-id: 69bb3cec0731ba325f60c6c8459426546c79b54d
Summary:
## Description
CompositeReactPackage will now eagerly initialize all NativeModules inside the TurboReactPackages. Before it would just crash the program because calling `createNativeModules()` on the TurboReactPackage instance. Longer term, we should make CompositeReactPackage a TurboReactPackage.
## Why am I making this change?
I made MainReactPackage into a TurboReactPackge. But ExpressWifiTechnician uses MainReactPackage to create a CompositeReactPackage:
**java/com/expresswifi/technician/MainActivity.java**
```
protected ReactPackage getPackage() {
ReactInstanceManager reactInstanceManager = getReactInstanceManager();
Nullable DevSupportManager devSupportManager = null;
if (reactInstanceManager != null) {
devSupportManager = reactInstanceManager.getDevSupportManager();
}
return new CompositeReactPackage(
super.getPackage(),
mXWFTechnicianReactPackage,
new RNFusedLocationPackage(),
mXWFTechnicialMobileConfigPackageProvider.get(devSupportManager));
}
```
**java/com/facebook/catalyst/shell/MainReactActivity.java**
```
public abstract class MainReactActivity extends AbstractReactActivity {
protected ReactPackage getPackage() {
return new MainReactPackage();
}
}
```
Reviewed By: fkgozali
Differential Revision: D15738677
fbshipit-source-id: 3220dfe6434de56f2917c77fb21acef4cfc79278
Summary: Removes `unsetOwner` from Yoga. This was temporarily for patching a crash.
Reviewed By: colriot
Differential Revision: D15737613
fbshipit-source-id: 8ab93ecf7ffb913df6207fe5db47a8cc93eded2c
