Summary:
X-link: https://github.com/facebook/yoga/pull/1382
Pull Request resolved: https://github.com/facebook/react-native/pull/39437
Have been running into places where C++ 20 makes life easier for use like `std::bit_cast` (that one is easy to polyfill), in-class member initializer support for bitfields, designated initializers, defaulted comparison operator, concepts instead of SFINAE, and probably more.
Our other infra is in the process of making this jump, or already has. This tests it out everywhere, across the various reference builds, to see if we have any issues.
This is a bit more aggressive than I had previously communicated, but n - 1 is going to be a better long term place than n - 2.
If we wanted to use `std::bit_cast` we would need one of:
1. GCC 11+ (~2.5 years old)
1. Clang 14 (~2.5 years old)
1. VS 16.11 (~2 years old)
For mobile this means:
1. NDK 26 (still in Beta 😭)
1. XCode 14.3.0 (~6 months old)
https://en.cppreference.com/w/cpp/compiler_support/20
That isn't quite doable yet, but we can start taking advantage of language features in the meantime. More of these will be supported in older toolchains.
Anyone needing support for older C++ versions can lag behind on more recent changes. E.g. Yoga 2.0 supports C++ 14.
bypass-github-export-checks
Changelog: [Internal]
Reviewed By: cortinico
Differential Revision: D49261607
fbshipit-source-id: ceb06eac20dfe93352d7b796d6847a7314069cf3
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39521
Minor Dev Menu changes:
- Improve guidance when failing to connect to the debugger.
- Align "Show Element Inspector" wording on iOS.
Changelog: [Internal]
Reviewed By: blakef
Differential Revision: D49375789
fbshipit-source-id: 092da6cae1a62a1bcc4dc877ab01369db6876741
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39495
We've put `[self unstable_registerLegacyComponents]` inside `#if RCT_NEW_ARCH_ENABLED` block, but we've missed the fact that it was inside the `else` clause of the `if (enableBridgeless)` statement. So it wasn't actually enabled for bridgeless mode regardless of `RCT_NEW_ARCH_ENABLED`.
This diff fixes that, and registers legacy components for the bridgeless mode as well.
Changelog: [Internal]
Reviewed By: NickGerleman, cipolleschi
Differential Revision: D49323913
fbshipit-source-id: 5b8e4ffc6ea1cdf5bbdecbddd5274d70c9b2eeab
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39502
This diff adds the parsing of the package.json to retrieve the version of react native and it matches it against a regex to decide whether to enable or not the New Architecture.
This change depends on [this CLI's PR](https://github.com/react-native-community/cli/pull/2076).
## Changelog:
[Android][Added] - Parse RN Version to decide whether to enable the New Arch or not.
Reviewed By: cortinico
Differential Revision: D49270012
fbshipit-source-id: 5164a1ddd26f1822c0d717a7a9e2416ec38d78a2
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39501
This change introduce a new model to parse the React Native package.json.
The objects are not connected but they will be used in the next diff
## Changelog:
[Android][Added] - Add a couple of types in the RNGP for json parsing.
Reviewed By: cortinico
Differential Revision: D49270044
fbshipit-source-id: 42a855bb85a5f8e1b1249ab3ebee974ce3fb93fd
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39500
When working on RNTeaster, we discovered that the Kotlin migration. broke something.
This diff make sure we can test RNTester
## Changelog:
[Internal] - Make RNTester's New Architecture Example execute again
Reviewed By: cortinico
Differential Revision: D49154282
fbshipit-source-id: cb1f6b076b129dc088cf2cbb5799b373b6afdcda
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39516
With Xcode15, Apple rewrote the C++ linker.
This is a breaking change that does not work with weak symbols.
As a workaround, apple is suggesting to add `-ld_classic` to the linker in order to readd support for weak symbols. The flag does not exists for Xcode 14.3 or lower, so we need to add it conditionally.
With this change, we introduce a couple of checks in the Hermes build logic:
1. Detect the version of Xcode that is used
2. Add the new flag to `HERMES_EXTRA_LINKER_FLAGS` if Xcode version is 15.
## Changelog:
[Internal] - Make hermes build properly with Xcode 15
Reviewed By: cortinico, dmytrorykun
Differential Revision: D49368675
fbshipit-source-id: 62d8ed81855c426f56ed94b6a2d6da2eb882b355
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39504
We are receiving "killed" signals in CircleCI for the analyze_pr jobs.
Looking at the doc, it seems that it is due to us allocating to much memory.
As a mitigation, let's bumpt the executor to Medium.
## Changelog:
[Internal] - Bump executors for Analyze-PR to Medium to avoid choking
## Facebook:
Small machines consume 5 cred/min. Medium consume 10.
On average, the job takes 7 min to complete. So we are paying 35 credits per job. We will be paying 70 creds.
The upper bound cost is $ 0.042 per execution.
In the past month we were running 1820 jobs. We will be spending 1820 * 0.042 = $76.44 per month, rather than $38.22.
This is roughly $500/year.
We can fix the script with a different approach in the future.
Reviewed By: cortinico, dmytrorykun
Differential Revision: D49368118
fbshipit-source-id: 7e04e4be79b0f07d02338834e543b707c93c4363
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39443
This task is unused in CI now and we can safely remove it from our build files.
Changelog:
[Internal] [Changed] - Remove unused downloadAll task
Reviewed By: mdvacca, cipolleschi
Differential Revision: D49233339
fbshipit-source-id: 558b44612b71bb1cb220cc21b8339835ae235302
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39481
This function is hardcoded to `true` and is never used, so I'm removing it.
I found no relevant usages in OSS so that can probably go.
Also see: https://github.com/facebook/react-native/issues/39292 where it got reported.
Changelog:
[Android] [Removed] - Remove unused `canLoadFile` function from HermesExecutor.java
Reviewed By: yungsters, mdvacca
Differential Revision: D49320244
fbshipit-source-id: e680aeb9a9e9e837c35b61eef43b3744ee68a726
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39474
When it comes to Xcode 15 RC, we are aware of two issues:
1. `unary_function` and `binary_function` not available in Cxx17
2. [Weak linking](https://developer.apple.com/documentation/xcode-release-notes/xcode-15-release-notes#Linking) is not supported anymore.
This change should fix both of the issues, adding the flags to allow for `unary_function`and `binary_function` to be called and adding the `-Wl -ld_classic` flag to `OTHER_LDFLAGS` in case Xcode 15 is detected.
## Changelog:
[Internal] - add the `_LIBCPP_ENABLE_CXX17_REMOVED_UNARY_BINARY_FUNCTION` and the `-Wl -ld_classic` flags to projects when needed
Reviewed By: dmytrorykun
Differential Revision: D49319256
fbshipit-source-id: bb895f1e60db915db79684f71fa436ce80b42111
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39478
When testing Xcode 15, we realized that a few pods we do not control directly have the IPHONEOS_DEPLOYMENT_TARGET set to old versions of iOS.
We can update that setting to silence the warning with Cocoapods and this is what this script does.
Notice that bumping that setting generated other warning as some APIs have been deprecated.
## Changelog:
[Internal] - Bump min IPHONEOS_DEPLOYMENT_TARGET for 3rd party pods
Reviewed By: dmytrorykun
Differential Revision: D49274837
fbshipit-source-id: 584d105c76d654daa2ecf5eb2f1b9381e70f567a
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39442
changelog: [internal]
Delete folly:small_vector. folly:small_vector was not used on Android, fragmenting platforms.
Additionally, we want to remove dependency on Folly from React Native.
Reviewed By: NickGerleman, rshest
Differential Revision: D48966287
fbshipit-source-id: 000b4d58d54b6100aeac016af53b5e5f8cbee705
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39459
## Changelog
[Android][Deprecated] - ReactModuleInfo constructor with getConstants arg is deprecated
introducing a new constructor that doesn't use getConstants and updating the internal codebase to use it. deprecated the old one since it's been copypasta'd a lot in oss.
Reviewed By: cortinico
Differential Revision: D49262824
fbshipit-source-id: 1a4d36b3153d8ce0ee869b8635c94005fe7c9594
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39479
The goal of this diff is to create a feature flag that will be used to enable/disable Stable API for Turbo Module.
This featureflag will be used to experiment with the new API in fb4a
changelog: [intenral] internal
Reviewed By: arushikesarwani94
Differential Revision: D49212022
fbshipit-source-id: 4f4830f11ba18ad849dc8b35f4e178e938b03fc9
Summary:
in this diff, we backout D49335986 to fix the issue where we are losing context from decoding an encoded string all at once via `NSURL URLWithString....`. this fixes the category of issues in https://fb.workplace.com/groups/rn.support/posts/25183724564582821.
however, this will regress the original issue that this fixed (https://fb.workplace.com/groups/rn.support/posts/25129344753354136), where encoding square brackets in the URI is causing percents to be re-encoded, resulting in a mixed uri with mixed encoding, i.e. `%2522%5B%5D`. we will be fixing this by not serving decoding brackets in our php serializer which was originally be the follow-up fix, since that does not respect the RFC.
bypass-github-export-checks
Reviewed By: fkgozali
Differential Revision: D49336101
fbshipit-source-id: 2012390b237de8a16a5e60e880905cec48584873
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39456
**Fixes new debugger launch flow on Android:**
D49158227 aimed to improve proxy-safe behaviour for remote dev servers by auto-detecting the appropriate server URL for clients using the `Host` header (etc) from the HTTP request. However, this approach broke the local case for Android emulators and externally connected devices since they would originate from a device-relative server hostname — e.g. `10.0.2.2` for the stock Android emulator.
https://pxl.cl/3mVmR
This commit reverts to an explicit approach where callers specify the base URL to the dev server that should be addressible from the development machine — now as a single `serverBaseUrl` option.
**Changes**
- Adds new `serverBaseUrl` option to `createDevMiddleware`, designed to be the base URL value for constructing dev server URLs returned in endpoints such as `/json/list`.
- This changes little for the `localhost` case (now enabling `https://` URLs), but enables remote dev server setups to specify this cleanly.
- Updates call site in `community-cli-plugin`.
Changelog: [Internal]
Reviewed By: robhogan
Differential Revision: D49276125
fbshipit-source-id: 2b6a8507073649832993971aa9d0870f54c9bd44
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39432
For the removal of JSI Module adding the `UIManagerProvider` interface as this will replace the `JSIModuleProvider`
Adding the `Deprecated` annotation since this would be deprecated later as part of Stable APIs with bridge removal and not encouraged usage.
Changelog:
[Internal] internal
Reviewed By: cortinico
Differential Revision: D49257901
fbshipit-source-id: 176ae03368309b56f7a7608419f1776baf4b512a
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39477
There is a typo inside `DefaultNewArchitectureEntryPoint` that we want to fix before this becomes part of the public API.
Changelog:
[Internal] [Changed] - Fix typo inside DefaultNewArchitectureEntryPoint
Reviewed By: sammy-SC, cipolleschi
Differential Revision: D49320026
fbshipit-source-id: 8cf4d2027860b8d2fafd2b8abc51e68392e6343e
Summary:
Update comment in RCTBundleURLProvider.h to use correct `jsBundleURLForBundleRoot` import with `fallbackExtension` instead `fallbackResource`
## Changelog:
- Update comment in `jsBundleURLForBundleRoot` definition on RCTBundleURLProvider.h file.
Pick one each for the category and type tags:
[IOS][CHANGED] - Updated comment in RCTBundleURLProvider.h to instance correct jsBundleURLForBundleRoot:fallbackExtension method.
Pull Request resolved: https://github.com/facebook/react-native/pull/39265
Test Plan: N/A
Reviewed By: cortinico
Differential Revision: D49227104
Pulled By: GijsWeterings
fbshipit-source-id: 6ccfb984ab6c1e858f1c568b2dd8f4626e133ffb
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39464
We currently have 854 warnings emitted due to nested namespaces being a C++ 17 extension lol.
This sets the standard explicitly to C++ 20 in the specs emitting the warnings (mostly RCT-Text).
It is undesirable that we add hundreds of warnings to the CocoaPods build without signal. Our Buck build and OSS Android build both treat warnings as errors. But... that is for another day.
Changelog: [Intenrnal]
Reviewed By: cortinico
Differential Revision: D49303081
fbshipit-source-id: 1233e9d51426fb8dbd8ca39a771e2fef3d0df856
Summary:
This is a reopened version of https://github.com/facebook/react-native/issues/35224 by isidoro98 which was closed without explanation, updated to resolve new merge conflicts and now includes an example in the RN-Tester app. Aside from that it is unchanged. Here is isidoro98's description from their original PR:
This PR builds on top of https://github.com/facebook/react-native/issues/31402, which introduced the `automaticallyAdjustsScrollIndicatorInsets` functionality. It aims to fix one of RN's longstanding pain point regarding the keyboard.
The changes provide a better way of handling the `ScrollView` offset when a keyboard opens. Currently, when a keyboard opens we apply an **offset** to the `Scrollview` that matches the size of the keyboard. This approach is great if we are using an `InputAccessoryView` but if we have multiple `TextInputs` in a `ScrollView`; offsetting the content by the size of the keyboard doesn't yield the best user experience.
## Changelog:
[iOS] [Changed] - Scroll `ScrollView` text fields into view with `automaticallyAdjustsScrollIndicatorInsets`
Pull Request resolved: https://github.com/facebook/react-native/pull/37766
Test Plan:
The videos below compare the current and proposed behaviors for the following code:
```js
<ScrollView
automaticallyAdjustKeyboardInsets
keyboardDismissMode="interactive">
{[...Array(10).keys()].map(item => (
<CustomTextInput placeholder={item.toString()} key={item} />
))}
</ScrollView>
```
| Current behaviour | Proposal |
|-|-|
|  |  |
As can be seen in the video, the **current behavior** applies an offset to the `ScrollView` content regardless of where the `TextInput` sits on the screen.
The proposal checks if the `TextInput` will be covered by the keyboard, and only then applies an offset. The offset applied is not the full size of the keyboard but instead only the required amount so that the `TextInput` is a **specific** distance above the top of the keyboard (customizable using the new `bottomKeyboardOffset` prop). This achieves a less "jumpy" experience for the user.
The proposal doesn't change the behavior of the `ScrollView` offset when an `InputAccessory` view is used, since it checks if the `TextField` that triggered the keyboard is a **descendant** of the `ScrollView` or not.
## Why not use other existing solutions?
RN ecosystem offers other alternatives for dealing with a keyboard inside a ScrollView, such as a `KeyboardAvoidingView` or using third party libraries like `react-native-keyboard-aware-scroll-view`. But as shown in the recordings below, these solutions don't provide the smoothness or behavior that can be achieved with `automaticallyAdjustsScrollIndicatorInsets`.
| KeyboardAvoidingView | rn-keyboard-aware-scroll-view |
|-|-|
|  |  |
As shown in the videos, the `TextInput` is hidden by the keyboard for a split second before becoming visible.
Code for the videos above:
```js
// KeyboardAvoidingView
<KeyboardAvoidingView
style={{flex: 1, flexDirection: 'column', justifyContent: 'center'}}
behavior="padding"
enabled>
<ScrollView>
{[...Array(10).keys()].map(item => (
<CustomTextInput placeholder={item.toString()} key={item} />
))}
</ScrollView>
</KeyboardAvoidingView>
```
```js
// rn-keyboard-aware-scroll-view
<KeyboardAwareScrollView>
{[...Array(10).keys()].map(item => (
<CustomTextInput placeholder={item.toString()} key={item} />
))}
</KeyboardAwareScrollView>
```
Reviewed By: sammy-SC
Differential Revision: D49269426
Pulled By: javache
fbshipit-source-id: 6ec2e7b45f6854dd34b9dbb06ab77053b6419733
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39454
1. Build with C++ 20 everywhere.
2. Update folly in OSS build to avoid a warning, and let us opt out of heavy bits required for supporting coroutines (we are still more than a year behind).
3. Update the folly podspec for the new version of folly
4. Update the many, many, copy/pasted header search paths for folly dependencies to be able to resolve fmt, now exposed to folly headers
4. Replaces a deprecated atomic usage with equivalent.
5. Fixes a deprecated implicit capturing `this` by value in lambda (same behavior, added a TODO because it seems sus)
Changelog: [Internal]
Reviewed By: sammy-SC
Differential Revision: D49271685
fbshipit-source-id: 16610f7bcd1032d5796cc11f1cfa92a33d8c2593
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39462
In this diff I'm introdcuing the new UnstableAPI called LazyTurboModuleManagerDelegate.
The goal of this class is to provide a Lazy registration of TurboModules without relying on ReactModule annotations neither the TurboReactPackage#getReactModuleInfoProvider method.
This is not the final implementation, it's just a prototype that will be used to experiment and learn about perf implications.
This class will be used in the next diffs of the stack
changelog: [intenral] internal
Reviewed By: cortinico
Differential Revision: D49172555
fbshipit-source-id: 1609c523e6e72f2aa42ca7cc35e9170104a101db
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39461
In this diff I'm creating an empty constructor on ReactPackageTurboModuleManagerDelegate. This will be used in the next diffs of the stack.
I'm also resorting methods for better read of the class
changelog: [intenral] internal
Reviewed By: arushikesarwani94
Differential Revision: D49172554
fbshipit-source-id: 54073c86c87aa145b12e783d8d8f75cd1f5b228c
Summary:
X-link: https://github.com/facebook/yoga/pull/1384
Pull Request resolved: https://github.com/facebook/react-native/pull/39446
This adds logic to the enum generator to generate C++ style scoped enums.
This gives us a few nicities over C enums, even if both must exist:
1. We can add types and keep unsgined enums directly in bitfields
2. Style/readability
3. Avoiding implicit int conversion
Reviewed By: rozele
Differential Revision: D49267996
fbshipit-source-id: 1c41164c377b317c1fef97811c46cbc00b5a837e
Summary:
X-link: https://github.com/facebook/yoga/pull/1380
Pull Request resolved: https://github.com/facebook/react-native/pull/39433
Back when rolling out flex gap, we encountered a bug where gap was added to the end of the main axis when a size was not specified.
During flex line justification/sizing, we calculate the amount of space that should be in between children. We erroneously add this, even after the last child element.
For `justify-content`, this space between children is derived from free space along the axis. The only time we have free space is if we had a dimension/dimension constraint already set on the parent. In this case, the extra space added to the end of the flex line is usually never noticed, because we bound `maxLineMainDim` to container dimension constraints at the end of layout, and the error doesn't effect how any children are positioned or sized.
There was at least one screenshot test where this issue showed up though, and I was able to add a slightly different repro where we may have free space without a definite dimension by enforcing a min dimension and not stretching.
{F1091401183}
The new reference is correct, and looking back at diffs, is what this seemed to originally look like when added three years ago. Seems like there may have been a potential regression, but I didn't spot anything suspicious when I looked around the code history.
`betweenMainDim` may still be set for `gap` even if we don't have a sized parent, which makes the extra space propagated to `maxLineMainDim` effect parent size.
Because we were in a code freeze, I opted to have us go with a solution just effecting flex gap, instead of the right one, in case there were any side effects. This cleans up the code to use the right calculation everywhere, and fixes a separate bug, where `endOfLineIndex` and `startOfLineIndex` may not be the last/first in the line if they are out of the layout flow (absolutely positioned, or display: none_
See the original conversation on https://github.com/facebook/yoga/pull/1188
Reviewed By: javache
Differential Revision: D49260049
fbshipit-source-id: 218552c5ff938668b9f257df7a1493e13ded4d0d
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/39458
## Changelog
[Android][Deprecated] - hasConstants in ReactModuleInfo is marked as deprecated
getting rid of `mHasConstants` ivar here and marking the getter as deprecated
Reviewed By: cortinico
Differential Revision: D49262577
fbshipit-source-id: 5394281bdc4a315ffc0d1debfaef1808dc6f4299
Nick Gerleman
Intl Scheduler
Alex Hunt
Rubén Norte
Dmitry Rykun
Samuel Susla
Riccardo Cipolleschi
Nicola Corti
Saad Najmi
Phillip Pan
David Vacca
Sunbreak
Arushi Kesarwani
Nate Stedman
Juliovaz
Adam Averay