Name already in use

Summary:
Pull Request resolved: #35109

This is just a safety check to make sure we don't accidentally
publish a 1.x or a 1000.x version on Maven Central by mistake.

Changelog:
[Internal] [Changed] - Gate the Maven Central publishing to 0.x version.

Reviewed By: mdvacca

Differential Revision: D40767782

fbshipit-source-id: 58f2906c3b01bfd0fd388a300ba303b289633d4e

…le.properties (#35108)

Summary:
Pull Request resolved: #35108

I've rewritten the comment in the android/app/build.gradle.
They were really old, contained wrong links and most of the people ignored it.

I've also moved the enabling of Hermes to the `gradle.properties` file.
RNGP still supports the old method, but we must be sure that we notify
library authors if they were reading project.ext.react.enableHermes in the past
(also the website needs to be updated).

I've also cleaned up the CircleCI setup as now we can specify Hermes enabled/disabled
via the CLI (this will also make easier to do e2e testing).

Changelog:
[Android] [Changed] - Cleanup the template documentation after RNGP & hermesEnabled to gradle.properties

Reviewed By: cipolleschi

Differential Revision: D40762872

fbshipit-source-id: 2c09245e0a923faac53cc6c8a89e99788ae47f8a

Summary:
Small PR with bump to the new versions of CLI and Metro in preparation of the branch cut for 0.71.

While at it, did a cheeky `npx yarn-deduplicate` to clean up a bit the deps.

## Changelog

<!-- Help reviewers and the release process by writing your own changelog entry. For an example, see:
https://reactnative.dev/contributing/changelogs-in-pull-requests
-->

[General] [Changed] - Bump CLI to 10.0.0-alpha.1 and Metro to 0.73.3

Pull Request resolved: #35107

Test Plan: CI is green

Reviewed By: motiz88

Differential Revision: D40762683

Pulled By: huntie

fbshipit-source-id: e523a49c78588ca80351f44cb02bcd4c0137475e

Summary:
Just bumping RNGP to make the new sources avialable to the template.

Changelog:
[Internal] [Changed] - Bump RNGP to 0.71.7

Reviewed By: cipolleschi

Differential Revision: D40760927

fbshipit-source-id: 909c88377a231ea6678c6af14c5594fdc4830b79

Summary:
`promise` module diff: [`8.2.0...8.3.0`](https://npmfs.com/compare/promise/8.2.0/8.3.0/)
- Hermes issue: facebook/hermes#766

## Changelog

[General] [Added] - Added support `Promise.any`

Pull Request resolved: #35080

Test Plan:
Release notes [`promise@8.3.0`](https://github.com/then/promise/releases/tag/8.3.0)

```tsx
typeof Promise.any // function
```

Reviewed By: cortinico

Differential Revision: D40681373

Pulled By: jacdebug

fbshipit-source-id: ecd589186483f3aa0f48da28a1f6dfcb1e26c8bc

Summary:
Pull Request resolved: #35095

This change will make sure that we load the correct JS engine at runtime,
by using the BuildConfig flag that RNGP sets for us.

This will solve a lot of noise in adb logcat for users seeing
stacktraces mentioning failing to load `jscexecutor` library.

This is also a breaking change, but as the API was not widely used nor
advertised in the template, we should be fine by just mentioning this in the release notes.

Changelog:
[Android] [Changed] - Update the template to load the correct JS engine at runtime

Reviewed By: cipolleschi

Differential Revision: D40710597

fbshipit-source-id: d59a7a52b22a9bf273ea89094c6620c3ecf6eb00

Summary:
Pull Request resolved: #35101

Those CMake variables are effectively unused now. They're raising a warning on CMake builds
of templates + let's not expose them as we haven't released RNGP yet, before libraries
or other tools start relying on them.

Changelog:
[Internal] [Changed] - RNGP - Do not set GENERATED_SRC_DIR and REACT_ANDROID_BUILD_DIR

Reviewed By: cipolleschi

Differential Revision: D40751998

fbshipit-source-id: 13f54a6247e4734c21c263f8b1e6b4b9e8ba406c

Summary:
Pull Request resolved: #35100

That's just a minor bump of AGP before the branch cut.

Changelog:
[Android] [Changed] - Bump AGP to 7.3.1

allow-large-files

Reviewed By: cipolleschi

Differential Revision: D40752006

fbshipit-source-id: 4856bc7ca275cf46d3afcc7c24928c5f1d5e6e33

Summary:
Pull Request resolved: #35099

I'm removing the ReactFlipperPlugin from the new app template.
That plugin is effectively empty and is a stub + is deprecated
so is generating a warning for all the users.

Changelog:
[Android] [Changed] - Do not import/use the deprecated ReactFlipperPlugin

Reviewed By: cipolleschi

Differential Revision: D40751984

fbshipit-source-id: 1f1fdf9c7dfccb9e7bbd4c8c2292df71e00c644f

Summary:
In 2017, React published v15.5 which extracted the built-in `prop types` to a separate package to reflect the fact that not everybody uses them. In 2018, React Native started to remove `PropTypes` from React Native for the same reason. In 0.68 React Native introduced a deprecation warning which notified users that the change was coming, and in 0.69 we removed the PropTypes entirely.

The feedback we've received from the community is that there has not been enough time to migrate libraries off of PropTypes. This has resulted in users needing to patch the React Native package `index.js` file directly to add back the PropTypes, instead of migrating off of them. We can empathize with this fix short term (it unblocks the upgrade) but long term this patch will cause users to miss important changes to `index.js`, and add a maintenance cost for users.

Part of the reason there was not enough time is that we didn't do a good job surfacing libraries that were using PropTypes. This means, when you got a deprecation warning, it wasn't clear where the source of the usage was (either in your code or in a library). So even if you wanted to migrate, it was difficult to know where to actually make the change.

In the next release, we've made it easier to find call sites using deprecated types by [fixing the code frame in errors](react-native-community/cli#1699) reporting in LogBox, and ensuring that [the app doesn't crash without a warning](#34650). This should make it easier to identify exactly where the deprecated usage is, so you can migrate it.

To help users get off of the patch, and allow more time to migrate, we're walking back the removal of PropTypes, and keeping it as a deprecation for a couple more versions. We ask that you either migrate off PropTypes to a type system like TypeScript, or migrate to the `deprecated-react-native-prop-types` package.

Once we feel more confident that the community has migrated and will not need to patch React Native in order to fix this issue, we'll remove the PropTypes again. **If you have any trouble finding the source of the PropType usage, please file an issue so we can help track it down with you.**

Changelog:
[General][Changed] - Add back deprecated PropTypes

Reviewed By: yungsters

Differential Revision: D40725705

fbshipit-source-id: 8ce61be30343827efd6dc89a012eeef0b6676deb

Summary:
## Overview

When I implemented `ignoreLogs` it was originally just to move `console.ignoreYellowBox` over to LogBox. When I did that, I also added filtering for console messages too. My thought process was: Developers probably don't want to configure hiding logs twice, so the LogBox method can handle both.

This was a mistake. We should never hide console errors and warnings, because it completely silences important feedback to the users such as deprecation warnings. These issues should be fixed, not silenced, and since adding the silencing behavior it's clear that this feature is being abused to ignore legitimate warnings that need address.

Issue #33557 is a great reason why - the correct fix for this is not to ignore the errors, it's to address the deprecation / removal of the API. Allowing an easy `ignoreLog` method to hide the problem made this migration much more painful.

Thus, we're reverting back to the pre-logbox behavior of always showing console logs, even if they're ignored by LogBox in the app UI. Hopefully, this results in more of these issue being addressed instead of ignored.

Changelog:
[General] [Changed] - Do not filter errors/warnings from console

Reviewed By: yungsters

Differential Revision: D40724661

fbshipit-source-id: de3d2db1b0c32dee96acf92c9b1ca07ba0f4e218

Summary:
We are running into a group seeing frequent disconnects from Metro in a specific office. These are surfaced (at least on iOS) as websocket closures, without a prior websocket error. WebSocket closure can be for a variety of reasons, and the spec for a CloseEvent is to include fields `wasClean`, `code`, and `reason`, with `code` having the most well-defined meaning.

This change makes it so that we emit extra context when the websocket is closed. That should help inform developers the reason behind any close that may be abnormal.

Changelog:
[General][Added] - Log Abnormal Closes to Metro Websocket

Reviewed By: motiz88

Differential Revision: D40660765

fbshipit-source-id: ef606d8d809af1c697a78eb00cc5666c29a8bca3

Summary:
Currently, RCTBlobManager (the native module for Blob support) cannot be loaded on iOS when the new architecture is enabled.

## Changelog

[General] [Added] - `BlobModule` to `RCTCoreModulesClassProvider`

Pull Request resolved: #35047

Test Plan:
The snippet below can be used to test Blob support with the new architecture enabled.

```
// App.tsx
import { useEffect } from 'react';
import { View } from 'react-native';
function uriToBlob(uri: any) {
  return new Promise((resolve, reject) => {
    const xhr = new XMLHttpRequest();
    xhr.responseType = 'blob';
    xhr.onload = () => {
      const blob = xhr.response;
      resolve(blob);
    };
    xhr.onerror = err => {
      reject(err);
    };
    xhr.open('GET', uri);
    xhr.send();
  });
}

export default function App() {
  useEffect(() => {
    uriToBlob('https://www.google.com/images/branding/googlelogo/1x/googlelogo_color_272x92dp.png');
  });
  return <View />;
}

```

Related issue: #35042

Reviewed By: NickGerleman

Differential Revision: D40716048

Pulled By: cipolleschi

fbshipit-source-id: 17643d230fa7ea83baee363d137d51f87818baa8

Summary:
This remove a flag to effectively go publishing to Sonatype.
The idea was to protect us against accidentally publishing a nightly as a stable release.
We need to remove this before RC0

Changelog:
[Internal] [Changed] - Remove the --dry-run gate from publishing to Sonatype

Reviewed By: cipolleschi

Differential Revision: D40687038

fbshipit-source-id: e6821905f41899430813f9575f17a5068b05a9bb

…p path (#35092)

Summary:
Pull Request resolved: #35092

As we're close to the cut of the 0.71 branch, I'll take the opportunity to polish
our Gradle public API.
We're exposing a mixture of Path (String) and File (java.io.File or RegularFileProperty).
Here I've moved everything to use File as it's more configurable for the users,
specifically if they're using monorepos or other setup.

This also allows us to remove the resolution logic for the cliPath.

Changelog:
[Internal] [Changed] - RNGP - Use the File Api to specify cliPath and remove ComposeSourceMap path

Reviewed By: cipolleschi

Differential Revision: D40710595

fbshipit-source-id: a17095eebae5123b70fd2b8e3d512656817006ca

…Architecture (#35091)

Summary:
Pull Request resolved: #35091

This diff refines the DefaultNewArchitectureEntryPoint to make it easier for user to
either turn on Fabric, TurboModules or both.

Changelog:
[Internal] [Changed] - Make it easier for user to toggle only Fabric or TurboModules in New Architecture

Reviewed By: cipolleschi

Differential Revision: D40710596

fbshipit-source-id: 236060b2ebccb1bf25e7f5c0fc15f54c5ce5f608

Differential Revision:
D40333083 (0fac981)

Original commit changeset: f3816e3bd7de

Original Phabricator Diff: D40333083 (0fac981)

fbshipit-source-id: 1a52a06b057c4c0ea6e3e4c3315ba73a883e3579

Summary:
`aria-checked` prop should accept `mixed` as value as given [here](https://www.w3.org/WAI/GL/wiki/Using_WAI-ARIA_aria-checked%3Dmixed) and also [accessibilityState.checked](https://reactnative.dev/docs/accessibility#accessibilitystate) accepts mixed to represent checkboxes. This change refers to issue #34424 and PR #34524

## Changelog

[General] [Added] - Added `mixed` value for `aria-checked`.

Pull Request resolved: #34633

Test Plan:
```js
<TouchableOpacity
  accessibilityRole="checkbox"
  aria-checked="mixed"
  accessibilityHint="click me to change state">
  <Text>Checkbox example</Text>
</TouchableOpacity>
```

Reviewed By: lunaleaps

Differential Revision: D39382158

Pulled By: necolas

fbshipit-source-id: fa026274111305cc0bcbb42ed974ca1be7d779a5

Summary: Changelog: [Internal] Flow InputAccessoryView

Reviewed By: yungsters

Differential Revision: D39855166

fbshipit-source-id: 9c389854e6d76a0c0e881190cd1808088de324fd

Summary:
Pull Request resolved: #35093

It turns out that using the Artifacts API to manipulate the APK to remove
.so has unintended side effects and is causing the `installDebug` and `installRelease`
commands to fail.

I've resorted to register a packaging option for each variant to make sure we include only
the correct artifacts we want.

This should fix the current startup crash that is experienced on main.

Changelog:
[Android] [Fixed] - RNGP - Do the .so cleanup using pickFirst and exclude

Reviewed By: cipolleschi

Differential Revision: D40722285

fbshipit-source-id: 982e1e9c474522fc4419c969ede5ee14e5404f3a

)

Summary:
Pull Request resolved: #35083

We don't need to `enableCodegenInApps` keys but we can instead rely on the
existence of the `codegenConfig` key inside the `package.json` to decide if
Codegen should run in App modules or not.

Changelog:
[Internal] [Changed] - RNGP - Remove enableCodegenInApps and infer it from package.json

Reviewed By: cipolleschi

Differential Revision: D40687079

fbshipit-source-id: cd4a6c67caa19c1d199ae75388a0551339f876a0

Summary:
Part of #34872
> [Assigned to youedd] This task is more advanced than the others Create a function to unify the [unionType](https://github.com/facebook/react-native/blob/main/packages/react-native-codegen/src/parsers/flow/modules/index.js#L372-L396) and the [TSUnionType](https://github.com/facebook/react-native/blob/main/packages/react-native-codegen/src/parsers/typescript/modules/index.js#L407-L431). This task may share some logic from the previous task. The function should accept a ParserType parameter to implement the proper logic to extract the memberType variable. Ideally, we should create a couple of supporting functions to implement those logics.

## Changelog

<!-- Help reviewers and the release process by writing your own changelog entry. For an example, see:
https://reactnative.dev/contributing/changelogs-in-pull-requests
-->

[Internal] [Changed] - Refactor translate UniontypeAnnotation

Pull Request resolved: #35050

Test Plan:
`yarn jest react-native-codegen`
![image](https://user-images.githubusercontent.com/19575877/197334084-2daeed7e-8e87-4140-b8c3-07f2de3f0496.png)

Reviewed By: cortinico

Differential Revision: D40637299

Pulled By: cipolleschi

fbshipit-source-id: 1490001a3398c3af79d465c40de3c3687f058523

Summary:
This PR is part of #34872.
This PR contains two changes, extracting the visitor object and then factoring out the `getConfigType()` function to the `parsers/utils.js` file. Then we can reuse the same function in both flow and typescript by passing the extracted visitor object.

## Changelog

[Internal] [Changed] - Extract visitor object in the same file and factor out `getConfigType()` to `parsers/utils.js`.

Pull Request resolved: #35035

Test Plan: Output of `yarn jest react-native-codegen` ensures all passed test cases

Reviewed By: cortinico

Differential Revision: D40548855

Pulled By: cipolleschi

fbshipit-source-id: 310b8565322a4e4800a3fffc67479a9dfa45d620

Summary:
changelog:
[general][Added] - Concurrent rendering safe implementation of ScrollViewStickyHeader

This is a re-land of ScrollViewStickyHeader from Kacie Bawiec.

Reviewed By: yungsters

Differential Revision: D40380217

fbshipit-source-id: 60dc86086a4d9d97eef71c4ed2e26536f7e72889

… warnings

Differential Revision:
D40716713 (8422c53)

Original commit changeset: eeb95a91c6cd

Original Phabricator Diff: D40716713 (8422c53)

fbshipit-source-id: e85f111e7da0381e09f8a23d3bd0d553b3a7c4a9

Summary:
I wanted to start working on a thing but this barrage of warnings was very annoying so ended up doing this instead: a very small PR to take care of some warnings during yarn install.

It doesn't change anything (the versions are the ones already used all around the repo), just makes yarn happier.

Also, while doing that I found a dependency that was lying there unused for 2 years so took care of it.

## Changelog

<!-- Help reviewers and the release process by writing your own changelog entry. For an example, see:
https://reactnative.dev/contributing/changelogs-in-pull-requests
-->

[General] [Fixed] - add wanted dependencies to remove yarn warnings

Pull Request resolved: #35088

Test Plan:
### Before

<img width="1920" alt="Screenshot 2022-10-26 at 10 53 32" src="https://user-images.githubusercontent.com/16104054/197996489-f463be29-b35b-45cc-9d9c-2d176579fb7d.png">

### After

<img width="947" alt="Screenshot 2022-10-26 at 10 52 19" src="https://user-images.githubusercontent.com/16104054/197996505-3d60b319-006b-45ab-83bf-2f431272fdcd.png">

Reviewed By: jacdebug

Differential Revision: D40716713

Pulled By: robhogan

fbshipit-source-id: eeb95a91c6cdf37edfe868fefe4ba04aebe500ec

)

Summary:
Pull Request resolved: #35089

Changelog:
[General][Fixed] - Back out "Add enum example to Android/iOS rn-tester TurboModule"

This broke the rn-tester adding due to an invalid flow-enum setup. Needs further investigation

Reviewed By: cipolleschi

Differential Revision: D40714320

fbshipit-source-id: 9831276762f90df0ffaca3304382fe5925009343

Summary:
This pull request migrates the activity indicator example to using React Hooks.

## Changelog
[General] [Changed] - RNTester: Migrate ActivityIndicator to hooks

Pull Request resolved: #35071

Test Plan: The animation works exactly as it did as when it was a class component

Reviewed By: jacdebug

Differential Revision: D40698379

Pulled By: NickGerleman

fbshipit-source-id: 08b275fcc4e3a10b5872e0031fa2ecce5360a7b9

Summary:
fd91748 added a condition not to pump an extra message when we didn't see symbolication warnings (which do not seem to show up in a stock internal build), but I missed that the test specific to RCTLogLevelError checked for string containment instead of equality, and the last assertion we process on log level is incorrect in OSS as a result.

Changelog:
[Internal][Fixed] - Fix RCTLoggingTests in OSS

Pull Request resolved: #35085

Test Plan:
`test_ios-Hermes` passes again in OSS: #35085 (JSC and Android failures look unrelated)

RCTLoggingTests still pass running `buck test //xplat/js/react-native-github/packages/rn-tester:RNTesterIntegrationTests`

Reviewed By: rshest

Differential Revision: D40699773

Pulled By: NickGerleman

fbshipit-source-id: 5acc8ec2b26a1f9acac2f070b85f1d65ee15a011

Summary:
* Add a DevToolsSettingsManager, which has android and iOS variants, which uses a new TM (Android) or takes advantage of the Settings TM (iOS) to get/set console patch settings
* This is backed by either the existing Settings module (iOS) or a new Java TM, which uses the SharedPreferences AP

## Testing

Manual testing

## Changelog

[General] [Added] - Add DevToolsSettingsManager

Pull Request resolved: #34964

Test Plan: * Extensive manual testing

Reviewed By: NickGerleman

Differential Revision: D40333083

Pulled By: rbalicki2

fbshipit-source-id: f3816e3bd7dea3086f6f2269c3a099af14aebb3b

Summary:
Without this, the CocoaPods output would display multiple "Building RNTester..." logs. With these changes, we make it clear that these are separate targets that are being configured.

Changelog: [internal]

Reviewed By: mdvacca, cipolleschi

Differential Revision: D40656813

fbshipit-source-id: a317112e804d08b9e64e2d62e8b27e4045b317b0

Summary:
Add enum example to Android/iOS rn-tester TurboModule

Changelog:
[General][Added] - Add enum example to Android/iOS rn-tester TurboModule

Reviewed By: javache

Differential Revision: D40619020

fbshipit-source-id: a113c5c78bcff3275e80d11bce8d0e6421b0b97d

Summary:
Some portions of continuous test runs RNTesterIntegrationTests will hang for an hour when running in Sandcastle without outputting anything. Locally debugging, I see a deadlock waiting on a semaphore, where the test assumed an extraneous symbolication warning would be fired.

This change makes it so that we only look for the extraneous messages if we haven't already seen the target message, since they do not seem to show up in local testing. I also added assertions per semaphore that the log statement is propagated within a still very gracious 10 seconds. This means we still can see what tests are running/failing if this test has issues, instead of a deadlocked test runner with no output.

Changelog:
[Internal][Fixed] - Fix deadlock in RCTLoggingTests

Reviewed By: cipolleschi

Differential Revision: D40667275

fbshipit-source-id: ab78f3c2ef47e2fd740b1dad2a65912e9542301d

Summary:
Pull Request resolved: #35079

This bumps the Docker image for React Native Android to 6.0
shipping the NDK 23 with it, finalizing all the work needed
to support NDK 23.

Changelog:
[Internal] [Changed] - Bump the Android Docker image to 6.0 (for NDK 23)

Reviewed By: cipolleschi

Differential Revision: D40675449

fbshipit-source-id: 5fb53080ce796263cd592dbc489743e6295060ba

Summary:
Pull Request resolved: #35075

This diff updates the New App template for Android to use the React Native Gradle Plugin.
With this we can:
1. Get rid of all the C++ code.
2. Remove a lot of New Architecture logic in the build.gradle
3. Reuse the prebuilts of React Native/Hermes via prefab

Changelog:
[Android] [Changed] - Update the template to use RNGP

Reviewed By: cipolleschi

Differential Revision: D40673732

fbshipit-source-id: 70935248993d1e24904c982e75f12ad580faa9d8