Summary:
Given async nature of RN, it is quite possible situation, hence it should not be a redbox.
And in Fabric it happends all the time which makes debugging painful.
Reviewed By: mdvacca
Differential Revision: D8473511
fbshipit-source-id: cfe7a1d3d105fde3b23db4c4c07b99864215c94c
Summary:
Enable CocoaPods test, iOS e2e.
Use parallelism to run several tests simultaneously within the same machine.
Circle CI
Closes https://github.com/facebook/react-native/pull/19764
Differential Revision: D8471955
Pulled By: hramos
fbshipit-source-id: c484fd6c66fb2d0d2305ced29e34cb305f73fb55
Summary:
Limit number of NDK jobs of ReactAndroid on CI using $BUILD_THREADS environment variable. Otherwise, it was spawning 32 jobs while building RNTester, which caused in OOM or unexpected failure.
CI: https://circleci.com/gh/dulmandakh/react-native/387
Android CI is green again 😍
Closes https://github.com/facebook/react-native/pull/19755
Differential Revision: D8466069
Pulled By: hramos
fbshipit-source-id: fd507d12c8e93f718e047719daa68aa25c37bfb2
Summary:
This makes JSTimers actively register a callback for callImmediates.
Besides being generally tricky, circular dependency prevent compiling React Native code with bundlers that rely on concatenating module sources rather that evaluating code at the time of requiring, like Google Closure Compiler.
Sadly, Google Closure Compiler setup that prompted this change is complicated and brittle. And there are no good public tools to find circular dependencies among Haste-style modules (with unqualified require paths).
So some advice on a good test plan would be useful. Does Facebook have any tools to find circular dependencies with Haste-style requires?
FWIW, a check that worked for me was to replace all import paths in React Native from Haste style to normal relative paths (which I needed anyway) and then run [`madge`](https://www.npmjs.com/package/madge) on the code base:
```
$ ~/node_modules/.bin/madge --circular react-native/Libraries
Processed 390 files (7.4s) (81 warnings)
✖ Found 2 circular dependencies!
1) BatchedBridge/NativeModules.js > BatchedBridge/BatchedBridge.js > BatchedBridge/MessageQueue.js > Core/Timers/JSTimers.js
2) StyleSheet/flattenStyle.js > StyleSheet/StyleSheet.js
```
(The second cycle is already eliminated in a8e3c7f578).
[GENERAL] [MINOR] [MessageQueue] - MessageQueue implementation doesn't have a circular dependency on JSTimers.
Closes https://github.com/facebook/react-native/pull/19526
Reviewed By: hramos
Differential Revision: D8458755
Pulled By: yungsters
fbshipit-source-id: e753139b920ba1ad1a6db10f974c03ca195340c7
Summary:
Suppress lint errors in Dialog module. remove abortOnError=false.
This might hide future problems, so removing it
``` gradle
lintOptions {
abortOnError false
}
```
Builds locally just fine. But CI is failing for unknown reasons. https://circleci.com/gh/dulmandakh/react-native/265
RNTester will built without errors
Closes https://github.com/facebook/react-native/pull/19740
Differential Revision: D8450600
Pulled By: hramos
fbshipit-source-id: faf508a0c546af18a05ee224628f88b02a38ab9f
Summary:
Android Support Library page says https://developer.android.com/topic/libraries/support-library/
> Caution: Starting with Support Library release 26.0.0 (July 2017), the minimum supported API level across most support libraries has increased to Android 4.0 (API level 14) for most library packages. For more information, see Version Support and Package Names in this document.
_android.support.v4.app.FragmentActivity_ is used to support **Fragment**s on Android API versions below 11. Now, we support only API version 14 and above, it's ok to remove _ReactFragmentActivity_ that extends _FragmentActivity_.
Once ReactFragmentActivity removed, we can remove some codes that use _android.support.v4_ to work support _ReactFragmentActivity_ on **unsupported** Android versions.
CI is failing for unknown reasons: https://circleci.com/gh/dulmandakh/react-native/278
> Received 'killed' signal
Everything will build and work just fine, except showing _ ReactFragmentActivity_ as deprecated
Closes https://github.com/facebook/react-native/pull/19741
Differential Revision: D8454968
Pulled By: hramos
fbshipit-source-id: e5f901438ef764163af013fe854904a28c73070a
Summary:
New metro releases are published to npm when a "Bump metro" commit is synced to GitHub and picked up by facebook/metro's Circle CI.
The same commit is picked up by facebook/react-native's Circle CI, but it will fail the checkout job because yarn will not find the new package on npm.
This is expected because the package will only be published to npm after Circle CI is done running on facebook/metro.
We can safely skip `yarn install` on these commits, and rely on the restored yarn cache instead.
Closes https://github.com/facebook/react-native/pull/19680
Differential Revision: D8450638
Pulled By: hramos
fbshipit-source-id: 436c229d09bfb4f9f92236c5aeec4a60d8317922
Summary: After moving all matrix math to C++, the actual client native code is quite trivial.
Reviewed By: fkgozali
Differential Revision: D8344059
fbshipit-source-id: 6910c6af5de64d5f901e82075d30edbde177af40
Summary:
This is quite interesting.
In Fabric, on C++ layer we store `color` values as `SharedColor` objects which (on iOS) are `shared_ptr`s with custom deallocater (which calls CoreFramework's `CFRelease` function). All this means that we cannot directly transfer ovenership of this managed pointer to ARC (honestly, I am not 100% sure about that, but at least this "shared ownershipt between ARC and non-ARC worlds" concept is as dangerous as any "relying on implementation details" approach). To to so, we have to create a copy and transfer ownership of the copy to ARC (which we do in 1RCTCGColorRefFromSharedColor`).
Reviewed By: fkgozali
Differential Revision: D8344061
fbshipit-source-id: 8b6764e1539b1982b41f502bbd3307c7b6900fd9
Summary: The matrix magic and parsing approach are mixins between current iOS and Android implementation.
Reviewed By: fkgozali
Differential Revision: D8344054
fbshipit-source-id: 524b48c5ab61959ce740373534d0d435eb37b647
Summary:
Trivial.
`nativeId` is meant to be used by subclasses of external inspector-like tools, so it does not have any real functionality.
Reviewed By: fkgozali
Differential Revision: D8344064
fbshipit-source-id: d86ef378cda1f7f0e9a7b4ffc09f51004ae530d2
Summary:
Just definition; we don't have an implementation on the native view layer yet.
And we don't have `transform` prop yet (because it's quite complex).
Reviewed By: fkgozali
Differential Revision: D8344058
fbshipit-source-id: 3b7b41480be8295cbc90b95ebe8562e52c6f81d7
Summary:
This is pretty straightforward implementation uses native `UISwitch`.
Suddenly we need Switch to test a bunch of other things.
Reviewed By: fkgozali
Differential Revision: D8344055
fbshipit-source-id: cfc51b8bc11198eb9d4d5e4745b96fb3a7f14de1
Summary:
This is just the convenient way to embed native views inside custom View components.
See coming <Switch> implementation as an example.
Reviewed By: fkgozali
Differential Revision: D8344056
fbshipit-source-id: 7f5f8cfeeffa7676bc7b562aa07f006cb9006575
Summary: CornerInsets is something like EdgeInsets but about corners instead of edges.
Reviewed By: fkgozali
Differential Revision: D8344062
fbshipit-source-id: 9bf7a8696fba96e3124cb15e8e84093c1f4f8747
Summary:
On JS reload the FabricUIManager and EventDispatcher didn't get release due to a retain cycle. This breaks the cycle.
In addition, force release the Scheduler on reload so that the stale classes get cleaned up properly, avoiding crashes. Also the surface now remounts the content correctly
Reviewed By: shergin
Differential Revision: D8414916
fbshipit-source-id: 4b14031f29b3bc9987d7aa765dc0d930a7de2b1e
Summary: Calling the event emitters on the main thread seems to be problematic, so let's dispatch it via the JS thread. This requires some changes to make "eventTarget" single-use because otherwise the binding would need to synchronize the actual JS call with the act of releasing the target.
Reviewed By: shergin
Differential Revision: D8375291
fbshipit-source-id: bd2b42731176ae209f4a19c232309c163fb1c01b
Summary:
Follow up to #18470, which has not yet landed in a release so we may want to hold off merging this until it has so that people receive the deprecation warning.
***
This PR removes the previously deprecated behavior of touchableHandleActivePressIn/Out (as well as onPressIn/Out being called on TV platforms for focus/blur. Instead, users should utilize the new `onFocus` and `onBlur` for these events.
This was because on TV platforms onPressIn/Out was overloaded to trigger for not only presses but these focus events as well. The normal behavior of true presses will still defer to touchableHandleActivePressIn/Out (which defers to onPressIn/Out).
<!--
Required.
Help reviewers and the release process by writing your own release notes. See below for an example.
-->
[TV] [BREAKING] [Touchable] - On TV platforms `onPressIn` and `onPressOut` on Touchables will now only be triggered for press events, not for focus/blur. Instead, you can use `onFocus` and `onBlur`.
<!--
**INTERNAL and MINOR tagged notes will not be included in the next version's final release notes.**
CATEGORY
[----------] TYPE
[ CLI ] [-------------] LOCATION
[ DOCS ] [ BREAKING ] [-------------]
[ GENERAL ] [ BUGFIX ] [ {Component} ]
[ INTERNAL ] [ ENHANCEMENT ] [ {Filename} ]
[ IOS ] [ FEATURE ] [ {Directory} ] |-----------|
[ ANDROID ] [ MINOR ] [ {Framework} ] - | {Message} |
[----------] [-------------] [-------------] |-----------|
EXAMPLES:
[IOS] [BREAKING] [FlatList] - Change a thing that breaks other things
[ANDROID] [BUGFIX] [TextInput] - Did a thing to TextInput
[CLI] [FEATURE] [local-cli/info/info.js] - CLI easier to do things with
[DOCS] [BUGFIX] [GettingStarted.md] - Accidentally a thing/word
[GENERAL] [ENHANCEMENT] [Yoga] - Added new yoga thing/position
[INTERNAL] [FEATURE] [./scripts] - Added thing to script that nobody will see
-->
Cc/ matthargett
Closes https://github.com/facebook/react-native/pull/19718
Differential Revision: D8450514
Pulled By: hramos
fbshipit-source-id: 818e85a338a451834b54d8965602699fc9b24e87
Summary:
Potential fixes for 18890. The issue was that when the setAttributedText method in RCTBaseTextInputView.m is called it does two main things:
Check that the attributed text that it receives is equal to the existing attributed text in the underlying backed up view ( backedTextInputView)
If not set the attributed text of the backed up view to the value passed into the method. This kills the dictation as it is effectively the equivalent of typing in the backed up text view.
self.backedTextInputView.attributedText = attributedText;
is the problem. It kills the dictation. It may have other effects as well.
In all cases I have seen the underlying text of the attributed string that is passed in the same as the text in the backedTextInputView, what was said to the dictation; however the attributes are different, which causes the isEqualToAttributedString: check to fail and thus the update happens, and the keyboard is killed.
Fix is to test for the underlying string equality not the attributed string equality when the input mode is dictation.
By necessity this had to be an integration test on an existing application. To test I enabled the keyboard and started the dictation. It worked with this fixes and not without. Will upload videos later.
This might break attributes on dictation, as it is happening. However anything set on the existing underlying text view should hold.
[IOS] [BUG] [Textinput] fixed an issue where the keyboard dictation ended abruptly.
Closes https://github.com/facebook/react-native/pull/19687
Differential Revision: D8450590
Pulled By: hramos
fbshipit-source-id: f97084131f98e9e0ed1f32111afc0f9f510f3b3b
Summary:
On pre-M devices, `PermissionsAndroid.request` currently returns a boolean, whereas on newer, it returns GRANTED, DENIED or other constants defined in `PermissionsModule.java`
given the example form the [docs](https://facebook.github.io/react-native/docs/permissionsandroid.html) which I guess many people use, this will lead to code that does not work before M (it will tell you that permissions are not given even if they are in the manifest).
I believe the author of [this](51efaab120) forgot to change the resolved value in this one place but changed it in other places, eg [here](51efaab120 (diff-2a74096453bc8faa5d4a1599ad0ab33fL99)).
The docs are written correctly:
> On devices before SDK version 23, the permissions are automatically granted if they appear in the manifest, so check and request should always be true.
but the code is not right because we'd need to check for `if (granted === PermissionsAndroid.RESULTS.GRANTED || granted === true) {`
Also, the behavior is done correctly in [requestMultiplePermissions](26684cf3ad/ReactAndroid/src/main/java/com/facebook/react/modules/permissions/PermissionsModule.java (L148)) so returning a boolean is an inconsistency.
I tested this locally. The code is the same as on line [148](26684cf3ad/ReactAndroid/src/main/java/com/facebook/react/modules/permissions/PermissionsModule.java (L148)) where it is done correctly.
[ANDROID] [BUGFIX] [PermissionAndroid] - return GRANTED / DENIED instead of true / false on pre-M android
<!--
**INTERNAL and MINOR tagged notes will not be included in the next version's final release notes.**
CATEGORY
[----------] TYPE
[ CLI ] [-------------] LOCATION
[ DOCS ] [ BREAKING ] [-------------]
[ GENERAL ] [ BUGFIX ] [ {Component} ]
[ INTERNAL ] [ ENHANCEMENT ] [ {Filename} ]
[ IOS ] [ FEATURE ] [ {Directory} ] |-----------|
[ ANDROID ] [ MINOR ] [ {Framework} ] - | {Message} |
[----------] [-------------] [-------------] |-----------|
EXAMPLES:
[IOS] [BREAKING] [FlatList] - Change a thing that breaks other things
[ANDROID] [BUGFIX] [TextInput] - Did a thing to TextInput
[CLI] [FEATURE] [local-cli/info/info.js] - CLI easier to do things with
[DOCS] [BUGFIX] [GettingStarted.md] - Accidentally a thing/word
[GENERAL] [ENHANCEMENT] [Yoga] - Added new yoga thing/position
[INTERNAL] [FEATURE] [./scripts] - Added thing to script that nobody will see
-->
Closes https://github.com/facebook/react-native/pull/19734
Differential Revision: D8450402
Pulled By: hramos
fbshipit-source-id: 46b0b7f83f81d817d60234f155d43de7f57248c7
Summary: The app needs to find the surface by rootTag at some point.
Reviewed By: shergin
Differential Revision: D8391957
fbshipit-source-id: b0bb107b7be882071890afb46de17b50e7ee060d
Summary:
The bug #18374 was caused by the loose condition to execute `stringByReplacingCharactersInRange` in the method `textInputShouldChangeTextInRange` . As a result, `findMismatch` wrongly returning `true` which ends up the Backspace event being fired in another `textInputShouldChangeTextInRange` call in `textInputDidChange`.
<!--
Required: Write your motivation here.
If this PR fixes an issue, type "Fixes #issueNumber" to automatically close the issue when the PR is merged.
-->
1. Pass all the tests by `yarn run test`
2. Run the following code and type any text. (This code is brought from #18374. Thank you michalpetrov!!) And then verify that 'Backspace' events are not emitted after clearing text
and entering any letters.
```javascript
type Props = {};
type State = {
text: string,
keys: string
};
export default class App extends Component<Props, State> {
state = {text: '', keys: ''}
render() {
return (
<View style={styles.container}>
<TextInput style={styles.textInput} value={this.state.text} onChangeText={this.onChangeText} onKeyPress={this.onKeyPress}/>
<Button title="Clear" onPress={this.onClear}/>
<Text>Text: {this.state.text}</Text>
<Text>Keys: {this.state.keys}</Text>
</View>
);
}
onChangeText = (text: string) => {
this.setState({text})
}
onKeyPress = ({ nativeEvent }: Object) => {
this.setState({keys: this.state.keys + nativeEvent.key + ', '})
}
onClear = () => {
this.setState({text: '', keys: ''})
}
}
```
<!--
Does this PR require a documentation change?
Create a PR at https://github.com/facebook/react-native-website and add a link to it here.
-->
<!--
Required.
Help reviewers and the release process by writing your own release notes. See below for an example.
-->
[IOS] [BUGFIX] [TextInput] - Fixed the bug where Backspace event was emitted when entering a character after clearing a text in TextInput by an empty string
<!--
**INTERNAL and MINOR tagged notes will not be included in the next version's final release notes.**
CATEGORY
[----------] TYPE
[ CLI ] [-------------] LOCATION
[ DOCS ] [ BREAKING ] [-------------]
[ GENERAL ] [ BUGFIX ] [ {Component} ]
[ INTERNAL ] [ ENHANCEMENT ] [ {Filename} ]
[ IOS ] [ FEATURE ] [ {Directory} ] |-----------|
[ ANDROID ] [ MINOR ] [ {Framework} ] - | {Message} |
[----------] [-------------] [-------------] |-----------|
EXAMPLES:
[IOS] [BREAKING] [FlatList] - Change a thing that breaks other things
[ANDROID] [BUGFIX] [TextInput] - Did a thing to TextInput
[CLI] [FEATURE] [local-cli/info/info.js] - CLI easier to do things with
[DOCS] [BUGFIX] [GettingStarted.md] - Accidentally a thing/word
[GENERAL] [ENHANCEMENT] [Yoga] - Added new yoga thing/position
[INTERNAL] [FEATURE] [./scripts] - Added thing to script that nobody will see
-->
Closes https://github.com/facebook/react-native/pull/18627
Differential Revision: D8436331
Pulled By: hramos
fbshipit-source-id: ec75a6ca926061cbf7cb106db652f2b4a71c9a0c
Summary:
* Current ci is missing an important part to test the whole part. With this we can make sure the js and android part compiles.
* Ensure the current android proguard rules is okay.
The `my-release-key.keystore` is just a copy of debug.keystore in `react-native/keystores`.
Pass all ci.
none
[GENERAL] [ENHANCEMENT] [CI] - Add RNTester to ci
Closes https://github.com/facebook/react-native/pull/19673
Differential Revision: D8435419
Pulled By: hramos
fbshipit-source-id: d3d92a5d1b8477c1f298643cc96695769e5c93ea
Summary:
Now that we have templates for everything we want to handle in the repo, the bot will go back to automatically closing issues that do not use one of the provided templates.
We have an escape valve for generic issues ("For Discussion"), so anyone who gets their issue closed automatically by the bot is unlikely to have read our instructions.
Closes https://github.com/facebook/react-native/pull/19723
Differential Revision: D8435415
Pulled By: hramos
fbshipit-source-id: 41db33cefce1367ad8f3d9440b7bba27565679cb
Summary:
bump gradle-plugin@2.3.3, gradle@3.5.1, gradle-download-task@3.4.3, as we landed build tools 26.0.2 with 065c5b6590
Will improve Android build performance.
Everything will work as normal, but build faster.
<!--
Required.
Help reviewers and the release process by writing your own release notes. See below for an example.
-->
[ANDROID] [ENHANCEMENT] [TOOLS] - bump gradle-plugin@2.3.3, gradle@3.5.1, gradle-download-task@3.4.3
Closes https://github.com/facebook/react-native/pull/19697
Differential Revision: D8433743
Pulled By: hramos
fbshipit-source-id: da72aeb314bed7f63807a0c69bebd24c633cc807
Summary:
This PR will add Google maven repo to RN project template that contains com.android.support:appcompat-v7:26.1.0, thus fixes `react-native run-android` using version 0.56.0-rc.1
CI is Green - https://circleci.com/gh/dulmandakh/react-native/235
new Android projects will build and run just fine.
Closes https://github.com/facebook/react-native/pull/19712
Differential Revision: D8433730
Pulled By: hramos
fbshipit-source-id: b7d5a1cd5a97b1c4aad2a307158d6dbfcf9a42a5
Summary:
Now that tests are green, we can return to automatic publishing to npm based on git tags. As long as all tests are passing, deploying a new release of React Native is as easy as tagging a commit. I've updated `Releases.md` to reflect today's release process.
Future Work: Include information about updating the website, as this is no longer done automatically.
Closes https://github.com/facebook/react-native/pull/19715
Differential Revision: D8429834
Pulled By: hramos
fbshipit-source-id: 2c6f2c80ac43c4e6d20c01e06ba14a7e4b16180d
Summary:
I met the error `Native module xyz tried to override xyz for module name xyzModuleName. If this was your intention...` after something went wrong during `react-native link` - one module somehow ended up being included twice in the `getPackages` method, as in:
```java
Override
protected List<ReactPackage> getPackages() {
return Arrays.<ReactPackage>asList(
new MainReactPackage(),
new WowPackage(),
new WowPackage(),
```
Since I have > 20 native modules it took me a little while to find out what the problem was. The improved error message should make the problem clearer to anybody who may encounter it. I did try to refactor the code a little more, by extracting the whole part of:
```java
String name = moduleHolder.getName();
if (namesToType.containsKey(name)) {
Class<? extends NativeModule> existingNativeModule = namesToType.get(name);
if (!moduleHolder.getCanOverrideExistingModule()) {
throw new IllegalStateException(getModuleOverridingExceptionMessage(
type.getSimpleName(),
existingNativeModule.getSimpleName(),
name
));
}
mModules.remove(existingNativeModule);
}
namesToType.put(name, type);
mModules.put(type, moduleHolder);
```
out into a separate method since there were two places where nearly identical code was written.
I have built RN from source and used it in a very simple app with RN vector icons (the package creates a native module as can be seen [here](https://github.com/oblador/react-native-vector-icons/blob/master/android/src/main/java/com/oblador/vectoricons/VectorIconsPackage.java#L19)).
After including the module twice, ie.
```java
Override
protected List<ReactPackage> getPackages() {
return Arrays.<ReactPackage>asList(
new MainReactPackage(),
new VectorIconsPackage(),
new VectorIconsPackage()
);
}
```
I get the improved error description, as seen in the screenshot.
<img src="https://user-images.githubusercontent.com/1566403/36340960-3289d9d0-13e7-11e8-8d17-e1651da17841.png" height="500">
[ANDROID] [MINOR] [NativeModuleRegistryBuilder] - Improve error message and refactor putting native modules to module maps
Closes https://github.com/facebook/react-native/pull/16402
Differential Revision: D8421392
Pulled By: hramos
fbshipit-source-id: 719bd37b4681933d35858621b402ae73dd460a5b
Valentin Shergin
David Vacca
Héctor Ramos
Dulmandakh
Sho Ikeda
Tim Yung
Kirill Nikolaev
Kevin Gozali
Héctor Ramos
Jay Phelps
Eoin Norris
Vojtech Novak
gengjiawen
Peter van der Zee
Eli White
hamaron
Jens Panneel
Miguel Jimenez Esun
