Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44083
Changelog: [Internal]
Adds a working, interactive "paused in debugger" overlay to Fusebox for iOS (similar to what D56068445 + D56098084 did on Android). The overlay is rendered using a standard `UIAlertController`. In upcoming diffs on this stack we'll style the overlay to look more like the Chrome implementation.
Reviewed By: hoxyq
Differential Revision: D56105959
fbshipit-source-id: d752d6611b2d9e48b67a82f3a9a96c7785c31a7a
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44114
In OSS, we should use the ClassFinder wrapper as the `Class.byName` calls will never succeed (we don't run the annotation processor in OSS).
Changelog:
[Internal] [Changed] - Use ClassFinder inside ViewManagerPropertyUpdater
Reviewed By: arushikesarwani94
Differential Revision: D56191175
fbshipit-source-id: a67195e983774872e27d35456b45651540411e2b
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44128
Changelog: [Internal]
idk what happened but this link got abbreviated when i copypastad it, fixing it here
Reviewed By: javache
Differential Revision: D56228228
fbshipit-source-id: 35c5c9fb44e82083773302618474c1e3bbc71712
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44080
## Design
Adds a new public `HostTarget::sendCommand` method, enabling integrators to send simple imperative commands to the target.
As an implementation detail, the commands are translated internally to CDP and sent over a dedicated `HostTargetSession` (encapsulated in `HostCommandSender`). Any response from the underlying Agent is ignored.
From the caller's perspective, these commands don't occur in the context of a session at all, and from the frontend's perspective, only the *effects* of the commands (if any) are seen.
## Use case
HostCommands are specifically useful when we want to resume/step execution in response to a UI action. The commands map directly to the `Debugger.resume` and `Debugger.stepOver` CDP methods.
NOTE: This is inspired by Chrome/V8's existing support for multiple concurrent CDP sessions. Any CDP client can successfully send `Debugger.resume` and `Debugger.stepOver` (without even subscribing to debugger events using `Debugger.enable`) and affect the state of other ongoing debugging sessions.
Changelog: [Internal]
Reviewed By: hoxyq
Differential Revision: D56098083
fbshipit-source-id: 013ab748b360f700c453cf1447fb82d6d0d77c6f
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44105
Previously, we skipped calling the ReactDelegate callback functions when the ReavtNavigationFragment was destroyed because it set the current activity to null when we need to actually keep the reference to the activity. However, skipping this entirely also skips core clean up logic, such as running the return() function for useEffect().
Instead of skipping the callback function entirely, we just need to make sure we don't set mCurrentActivity to null. I followed D30504616 to configure an option to not set mCurrentActivity to null if mKeepActivity flag is set on the ReactInstanceManager.
Changelog: [Internal]
Reviewed By: keoskate
Differential Revision: D56167533
fbshipit-source-id: cb3620e21599683e0c6bbc5a6a9c4f384fdbcc51
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44104
We previously added the option to skip calling the delegate lifecycle events in D55646221 to support fragment-based navigation. However, we don't actually want to skip calling these events as they run some core clean up logic, such as calling the return value of useEffect(). There's a better way to get this working with fragment-based nav (see next diff).
Changelog:
[Internal] [Changed] - Remove option to skip calling delegate on ReactFragment lifecycle events
Reviewed By: keoskate, cortinico
Differential Revision: D56167614
fbshipit-source-id: 3a4b91a303a27c0e19644a4e6611229211a1e530
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44120
changelog: [internal]
Prevent measuring the same text twice in `ParagraphShadowNode`.
The current implementation calls `TextLayoutManager::measure` twice for a single `ParagraphShadowNode`. The first time to measure the node for Yoga. The second time inside `ParagraphShadowNode::layout`. I think the original author counted on the cache inside of `TextLayoutManager` to deal with this, but this is not always the case and `TextLayoutManager::measure` is called with two different available widths, leading to cache miss and fills the cache faster.
Reviewed By: NickGerleman
Differential Revision: D55757264
fbshipit-source-id: 0bf8b49f062f802a4e2f04cad1bf1d4bf001b870
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44094
# Changelog: [Internal]
This script was added in D54770207.
This will be DEV-only for 2 reasons:
1. We need to double-check if Fusebox is ready to be used with production bundles, I don't think so.
2. Previous integration with RDT was DEV-only
Reviewed By: motiz88
Differential Revision: D56141041
fbshipit-source-id: 1141c65a5811d0f56c944e25a319ba21211b7836
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44088
This is a power user option for Release Crew members when testing locally — a flag to bypass being blocked on the latest in-progress CircleCI job and instead fetch build artifacts from the most recent successful pipeline (typically `HEAD~1`).
Example use cases where the latest pushed commit isn't impactful:
- An iOS-only fix, meaning Android can be tested now.
- A trivial fix that applies to CI only (e.g. RNTester Podfile.lock update).
Changelog: [Internal]
Reviewed By: cortinico
Differential Revision: D56138727
fbshipit-source-id: f9884bdb289a92486807e8e033b756466fcec559
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44116
This is just the bridgeless analogue to .hasCatalystInstance()
Changelog: [Android][Added] - Introduced ReactContext.hasReactInstance() to replace .hasCatalystInstance()
Reviewed By: fabriziocucci
Differential Revision: D56164488
fbshipit-source-id: 8be4676e18dc7df4765746f46cf36e62405b4ffa
Summary:
As part of decoupling our dependency on the react-native-community/cli, the `react-native.config.js` which is
a part of the community's config ecosystem should probably be entirely removed from the react-native package.
As part of this, we're making the config fail gracefully if these dependencies aren't available in a user's project:
- react-native-community/cli-platform-android
- react-native-community/cli-platform-ios
Changelog: [Internal]
This isn't going to be a visible change to any users.
bypass-github-export-checks
Reviewed By: cortinico
Differential Revision: D56137820
fbshipit-source-id: 528e25809a83b90e79b806a875001fc0f06db1cf
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44078
Changelog: [Internal]
Adds stub support for the [`Overlay.setPausedInDebuggerMessage`](https://cdpstatus.reactnative.dev/devtools-protocol/tot/Overlay#method-setPausedInDebuggerMessage) CDP method to `HostAgent` in the Fusebox backend, and propagates it into the Android and iOS integrations through `HostTargetDelegate`.
We take care to call `HostTargetDelegate::onSetPausedInDebuggerMessage()` a final time with a null `message` parameter, regardless of whether the client has actually sent the corresponding CDP message. Since multiple clients might be connected concurrently, we only send the `null` message when the *last* client which has requested a non-null message has disconnected.
Reviewed By: robhogan
Differential Revision: D56068444
fbshipit-source-id: c26e1cf17dec8d7dbb7edd5ab7fa3133642628ff
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44092
In a setup where a device retrieves a bundle from `http://127.0.0.1:8081`, but this is tunnelled to a remote host with only an IPv6 stack (eg, FB dev servers), the host running the inspector-proxy will fail to fetch source or source maps from 127.0.0.1 despite typically being on the same host (indeed, process) as Metro.
This causes a surprising inconsistency where using a bundler URL of `localhost` from the device results in source maps being inlined into `Debugger.scriptParsed`, but using a bundler URL of `127.0.0.1` causes inspector-proxy to fall back to preserving URLs, which are typically fetched lazily by CDT later.
This should be unnecessary once we've implemented CDP `Network.loadNetworkResource` and removed `Debugger.scriptParsed` rewriting, but for now it brings IPv6 tunnelled servers in line with local servers.
Changelog:
[General][Changed] Inspector proxy: Rewrite 127.0.0.1 to localhost in source map URLs for better IPv4->IPv6 tunnelling support.
Reviewed By: motiz88
Differential Revision: D56138742
fbshipit-source-id: b65c9cc8225a0ed54cf32171f640ef9e6408c762
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44103
Changes here fixes events in PopupMenuAndroid not being triggered correctly.
Issuses were:
1) naming mismatch
2) wrong parameters were set for the Event Map
3) missing code in .cpp
Applied fixes:
1) consistent event naming
2) fixed key used for event mapping
3) re-ran codegen to update .cpp files
## Changelog:
[Android] [internal] - Fix issue with PopupMenuAndroid event callback not working
Steps took to run codegen for this diff: https://www.internalfb.com/intern/phabricator/paste/markdown/P1214671854/
This diff is patching issues from D55531870
Reviewed By: RSNara
Differential Revision: D56164235
fbshipit-source-id: 4cf66ad3cfd753c146c5e219f27910834731e183
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44098
This sets `QualiferAlignment` so that code is automatically formatted to west const.
I did a pass at this before, but now that we are on new Clang Format, we can enforce it automatically, and I think a couple more cases not previously changed now are.
Changelog: [Internal]
Reviewed By: christophpurrer
Differential Revision: D56143678
fbshipit-source-id: 8f12b288476ea6019fd7d7a93a39b4fe2e75af14
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44089
While doing the release of 0.74.0-RC.9, we encountered some failures while building because the committed project had some new flags that have been added by cocoapods and that required Ccache.
However, Ccache is not a hard requirement to run React Native and the build started failing on systems that does not have ccache installed.
That happened because we were missing the piece of code that removed Ccache from the project in case the tool is not installed in the system.
We already committed such commit in the stable branch of 0.74, with [this commit](https://github.com/facebook/react-native/commit/2b18fdf8063b423a0fb5762f2c6044244b4c35e6). This change will port the same fix in main.
## Changelog
[iOS][Fixed] - Make sure to remove ccache scripts when ccache is not installed
Reviewed By: cortinico
Differential Revision: D56140015
fbshipit-source-id: 24e7ebb4e5c08766b29705e8b6f03c3f164a96ab
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44086
When a debugger frontend is connected to inspector-proxy via another proxy or tunnel that times out on idle (such as [VS Code's remote tunnel](https://github.com/microsoft/vscode/blob/main/src/vs/platform/tunnel/node/tunnelService.ts)), the connection between proxy and debugger may be dropped.
In addition, when the connection is dropped without a closing handshake, the proxy does *not* detect the disconnection - no disconnect is logged to the reporter and no notifications are sent to any connected devices.
This adds a mechanism using the WebSocket-standard `ping` and `pong` frames to:
1. Keep the connection alive
2. Detect when the debugger has gone away
Note that as all WebSocket clients already **must** reply to a ping with a pong, this is non-breaking for compliant implementations: https://datatracker.ietf.org/doc/html/rfc6455#section-5.5.2
Changelog:
[General][Added] Inspector proxy: Add ping/pong keepalive to debugger connections.
Reviewed By: hoxyq
Differential Revision: D56069185
fbshipit-source-id: e322de631c652a502f3d554c15ed5412a751ee04
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44087
The `mFabricEnabled` field is initialized to false. This is a misalignment with how other classes are behaving like
ReactActivityDelegate.
Changelog:
[Internal] [Changed] - ReactFeatureFlags.mFabricEnabled should default to ReactFeatureFlags.enableFabricRenderer
Reviewed By: arushikesarwani94
Differential Revision: D56013057
fbshipit-source-id: fcab903ab42d3b30094dcebbcf5b662cd2f2c506
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44040
This is just a clarification for the KDoc of this property.
Changelog:
[Internal] [Changed] - Clarify documentation for debuggableVariants
Reviewed By: andrewdacenko
Differential Revision: D56012825
fbshipit-source-id: 837d2dbc0f7ca5853ba1cedf71a4a5c36661318f
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/43722
This change adds a drawable, when when drawn on the bounds of a border-box sized view, will draw a spec compliant box-shadow outside the box. This is reliant on Android `RenderNode` and `RenderEffect` APIs provided by API 31.
Inset box shadows can also be added using a similar method, but this is not done yet.
The code which manages this is in flux, but the underlying drawable should be good. Will add some tests once it's more wired up.
Changelog: [Internal]
Reviewed By: javache, cortinico
Differential Revision: D55561465
fbshipit-source-id: 6180568cff2779b826e73bb9184dbe042863b262
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/43983
We duplicate some pretty hairy code related to conversion between logical and physical edges, along with the grafting between uniform and non uniform radii. This encapsulates border radius resolution/assignment logic.
Changelog: [Internal]
Reviewed By: alanleedev
Differential Revision: D55635743
fbshipit-source-id: 906c35af2bf18f0586d71d05f9cf61d4248ede1e
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/43721
Makes some changes to `FilterHelper` to expose RenderEffects for filters publicly to other classes in the module. We use this in box-shadow, in order to reuse logic for sigma accepting blur filters.
Also fixes a crash related to the conversions back and forward between sigma and radius, where small values would cause a crash.
Changelog: [Internal]
Reviewed By: joevilches
Differential Revision: D55563775
fbshipit-source-id: fabc888eecb451e75a88c8633fe5cca1f6644faa
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/43720
This lets us use BG drawing code in rules which view depends on. I also removed some lib usage.
The original class is still present, subclassing the class in its new location, but is marked deprecated.
Next diffs in stack clean up some of our own now deprecated usage.
Changelog:
[Android][Breaking] - Deprecate `ReactViewBackgroundDrawable` in favor of `CSSBackgroundDrawable`
Reviewed By: javache
Differential Revision: D55565035
fbshipit-source-id: 501b3e2f674c09a88b1825657ba6349823054e8c
Summary:
After a [recent change](https://github.com/facebook/react-native/commit/90296be1d4fab09a52e02dd09f34f819136d0a07) we break part of the integration with the debug menu, which is was using the presence/absence of the bridge to decide whether we were in bridge or bridgeless.
For backward compatibility reasosn, the bridge ivar is now populated with the bridgeProxy, so just checking whether is nil or not is not enough to verify whether we are in bridge or in bridgeless mode anymore.
## Changelog:
[iOS][Fixed] - Make sure that the Open Debugger appears in bridgeless mode
Reviewed By: fkgozali
Differential Revision: D56067897
fbshipit-source-id: e2501ed730ff35bc755c24ef400130c551032e28
Summary:
We would set the value of _bridge ivar to bridgeProxy for turbo module in bridgeless mode in https://github.com/facebook/react-native/issues/43757 , so we need to change the way of bridgeless/bridge check.
## Changelog:
[IOS] [FIXED] - Change bridgeless check in dev menu
Pull Request resolved: https://github.com/facebook/react-native/pull/43976
Test Plan: Dev menu shows bridgeless/bridge mode correctly.
Reviewed By: christophpurrer
Differential Revision: D56056640
Pulled By: cipolleschi
fbshipit-source-id: 1358c3027c1d5f12c70dd4486cc1d5975c7a185a
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44056
## Changelog:
[Internal] -
Was looking into how devsupport is exactly implemented on different platforms (in the context of doing it for a new platform) and figured I may just well convert this to Koltin in the process (helped to understand inner working details as well).
Reviewed By: christophpurrer
Differential Revision: D56058560
fbshipit-source-id: 1e2ffcec480c5fa3fd8b6494c29a0db6f94aec78
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44055
## Changelog:
[Internal] -
Was looking into how devsupport is exactly implemented on different platforms (in the context of doing it for a new platform) and figured I may just well convert this to Koltin in the process (helped to understand inner working details as well).
Reviewed By: christophpurrer
Differential Revision: D56052137
fbshipit-source-id: 25280a57f46bec95dc1437ea6eb2eef08332797f
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/44058
## Changelog:
[Internal] -
This was revealed when running an ASAN build on MacOS - we were doing an unsafe downcast from `LayoutableShadowNode->ViewShadowNode` inside `ShadowTree::emitLayoutEvents`.
Which, even though incidentally worked, is generally unsafe, as we may get e.g. `ImageShadowNode` there, which doesn't inherit from `ViewShadowNode`.
That downcast to `ViewShadowNode` wasn't even required, to begin with, as all the needed information can be already extracted from the `LayoutableShadowNode` itself.
Reviewed By: christophpurrer, javache
Differential Revision: D56062334
fbshipit-source-id: 08d5b3f5e0c57dc51b051d23506c7933581fea29
Summary:
The goal of this PR is to allow the usage of `RCTRootViewFactory` from Swift. The issue with `RCTTurboModuleManager.h` is that it uses C++ in its header file, which is not allowed in Swift, making this initializer unavailable.
This PR allows users to just pass configuration + adds a nullable annotation to bundleURL.
Example usage:
```swift
import Foundation
import UIKit
import React
import React_RCTAppDelegate
main
class AppDelegate: NSObject, UIApplicationDelegate {
var window: UIWindow?
private var rootViewFactory: RCTRootViewFactory?
func application(
_ application: UIApplication,
didFinishLaunchingWithOptions launchOptions: [UIApplication.LaunchOptionsKey : Any]? = nil
) -> Bool {
// Create config
let config = RCTRootViewFactoryConfiguration(
bundleURL: self.bundleURL(),
newArchEnabled: true,
turboModuleEnabled: true,
bridgelessEnabled: false
)
// Create rootview factory
rootViewFactory = RCTRootViewFactory(configuration: config)
// Create rootview
let rootView = rootViewFactory?.view(withModuleName: "RN0740RC4")
let rootViewController = UIViewController()
rootViewController.view = rootView
// Create window and assign view controller
window = UIWindow(frame: UIScreen.main.bounds)
window?.rootViewController = rootViewController;
window?.makeKeyAndVisible()
return true
}
func bundleURL() -> URL? {
RCTBundleURLProvider.sharedSettings().jsBundleURL(forBundleRoot: "index")
}
}
```
## Changelog:
[IOS] [FIXED] - Allow usage of `RCTRootViewFactory` from Swift
Pull Request resolved: https://github.com/facebook/react-native/pull/43590
Test Plan: CI Green, check usage of initializer without turbo module delegate
Reviewed By: christophpurrer
Differential Revision: D56055938
Pulled By: cipolleschi
fbshipit-source-id: c80d9f7f707c376f590f3dc4c9bb8f88f2e57e6a
Moti Zilberman
Nicola Corti
Phillip Pan
Jane Li
Samuel Susla
Ruslan Lesiutin
Alex Hunt
Ramanpreet Nara
Andrew Datsenko
zhongwuzw
Biki-das
Blake Friedman
Rob Hogan
Alan Lee
Soe Lynn
Nick Gerleman
Pieter De Baets
Riccardo Cipolleschi
Arushi Kesarwani
Andres Suarez
Sam Zhou
Ruslan Shestopalyuk
Oskar Kwaśniewski
Fabrizio Cucci
Dmitry Rykun