Summary:This change adds some basic unit tests for native animated traversal algorithm. The following tests are added:
1) Build simple animated nodes graph, verify that frame-based animation execute updates and when it runs out of the frames we no longer schedule updates for the native view
2) Build simple animated nodes graph and start short timing animation, verify that animation JS callback gets called.
As a part of this change I'm fixing an issue that tests allowed me to discover, where I forgot to clear updates queue at the end of `runUpdates` method. It was causing the view to be updated even if there was no active animation for it (actually it was mitigated by another bug in `hasActiveAnimations` I'm fixing here too).
I'm also adding Nullable annotation in a bunch of places. To lazy to send it as a separate change - sorry.
Going forward I'm planning on adding more tests. Currently the number of nodes is pretty limited so it's difficult to construct more complex graphs, but once I land Add/Multiply
Closes https://github.com/facebook/react-native/pull/6858
Differential Revision: D3168549
Pulled By: astreet
fb-gh-sync-id: 5295c75f3c7817775b5154bb808888650ff74e12
fbshipit-source-id: 5295c75f3c7817775b5154bb808888650ff74e12
Summary: Maybe a slight perf improvement. Also helps get rid of the 'ReferenceError: __fbBatchedBridge' error in favor of a more explicit "you're calling JS functions before the bundle is loaded"
Differential Revision: D3142966
fb-gh-sync-id: 6a50efe7d1634248107c8c95a3e014cb263a9ca5
fbshipit-source-id: 6a50efe7d1634248107c8c95a3e014cb263a9ca5
Summary:Hey there and thanks for submitting a pull request! Please have a look at the following checklist so that others have enough information to review your pull request:
**motivation**
WebSocket spec supports [Sec-WebSocket-Protocol](https://tools.ietf.org/html/rfc6455#section-11.3.4) as a standard way for negotiate a sub protocol between client and server.
* ios WebSocket implementation supports it.
* android WebSocket implementation ignores this header, leave a comment syas: "OkHttp will overrides it", so it did not implement.
* after some test, OkHttp doesn't override the header we add.
**Test plan (required)**
1. run and react-native app on android
2. at the main page, invoke: `var ws = new WebSocket('ws://example.ws-service.fakedomain.com', 'my-sub-protocol');`
3. see the header if it send the correct header, ex, use ngrep: `sudo ngrep -t -Wbyline -deth0 host example.ws-service.fakedomain.com and port 80`
you should see the WebSocket initial GET handshake includes header:
`Sec-WebSocke
Closes https://github.com/facebook/react-native/pull/6223
Differential Revision: D3162822
fb-gh-sync-id: a00f1c0f3e1c24ad6aa234329cbb2abad7664264
fbshipit-source-id: a00f1c0f3e1c24ad6aa234329cbb2abad7664264
Summary: Creates a signature of the method that can be used for efficiently doing things based on the argument types.
Reviewed By: astreet
Differential Revision: D3147620
fb-gh-sync-id: da1419b96d61f5fc40861625d816c18b3c19b425
fbshipit-source-id: da1419b96d61f5fc40861625d816c18b3c19b425
Summary:* Add ability to configure the app that should open when starting debugging
axemclion discussed this feature with tadeuzagallo and martinbigio on: https://github.com/facebook/react-native/issues/5051
Closes https://github.com/facebook/react-native/pull/5683
Reviewed By: martinbigio
Differential Revision: D2971497
Pulled By: mkonicek
fb-gh-sync-id: 91c3ce68feed989658124bb96cb61d03dd032599
fbshipit-source-id: 91c3ce68feed989658124bb96cb61d03dd032599
Summary: If a ProgressBar is created on the shadow thread (for measurement) at the same time one is created on the UI thread, we hit a race condition which sometimes results in a crash (##java.lang.NullPointerException: Null pointer exception during instruction 'invoke-static {v5}, android.animation.AnimatorSet$Node android.animation.AnimatorSet$Node.access$200(android.animation.AnimatorSet$Node) // method@153'##). This diff synchronizes the ctors.
Reviewed By: foghina
Differential Revision: D3151037
fb-gh-sync-id: 8ed96d6debdc1a5fa2f9bd037bfb75def25be3e8
fbshipit-source-id: 8ed96d6debdc1a5fa2f9bd037bfb75def25be3e8
Summary:Remove Trailing Spaces.
Why:
Sometimes there are conflicts with trailing spaces
Saves space
Those whose tools automatically delete them will have their pr watered down with trailing space removal
Closes https://github.com/facebook/react-native/pull/6787
Differential Revision: D3144704
fb-gh-sync-id: d8a62f115a3f8a8a49d5b07f56c540a02af38cf8
fbshipit-source-id: d8a62f115a3f8a8a49d5b07f56c540a02af38cf8
Summary:In the code that extracts and validates arguments from a `ReactMethod`, there is verification that if a method contains a Promise in it's list of arguments that it must come last. This fix makes sure that the `executorTokenOffset` is taken into account when asserting that condition.
Closes https://github.com/facebook/react-native/pull/6633
Differential Revision: D3143207
fb-gh-sync-id: ae9ebd9d829f88993f9951c4cb2452b3f7618476
fbshipit-source-id: ae9ebd9d829f88993f9951c4cb2452b3f7618476
Summary: We're seeing intermittent crashes in ~Bridge() where m_destroyed isn't set. This could be because the value of m_destroyed is cached for the destructing thread and doesn't see that the value got updated. Using an atomic boolean should fix this.
Reviewed By: mhorowitz
Differential Revision: D3126701
fb-gh-sync-id: 5887edef748cc05971765943de80187ab7fd8ede
fbshipit-source-id: 5887edef748cc05971765943de80187ab7fd8ede
Summary:Update to latest master version of css-layout. Update integration in RCTShadow(Root)View to match.
This solves the issue with items not strechting vertically in column layouts (https://github.com/facebook/css-layout/issues/127)
Reviewed By: vjeux
Differential Revision: D3120699
fb-gh-sync-id: beba162e1255d3527e1160e9bd414a712cb10713
fbshipit-source-id: beba162e1255d3527e1160e9bd414a712cb10713
Summary:Context #4658
I kept the original commit and author.
cc mkonicek
Closes https://github.com/facebook/react-native/pull/6639
Differential Revision: D3126336
Pulled By: mkonicek
fb-gh-sync-id: 5ae7b37f0eb1db355bb87076d621a405ff9c23c5
fbshipit-source-id: 5ae7b37f0eb1db355bb87076d621a405ff9c23c5
Summary:Creating views shouldn't have side effects so we can start doing them before the end of a batch. In order to limit the effect on framerate, we stop executing these create view operations after we've passed halfway through the frame.
Note, this doesn't seem to work yet for nodes: I'll address that in a followup diff.
Reviewed By: lexs
Differential Revision: D3120631
fb-gh-sync-id: 981540fac5a7499158146adb72c1be21fd0b5702
fbshipit-source-id: 981540fac5a7499158146adb72c1be21fd0b5702
Summary: We lost java module and method names in systrace when we moved to the xplat bridge.
Reviewed By: mhorowitz
Differential Revision: D3116042
fb-gh-sync-id: a61181f4d6be04c70122b7f79d426e364df1e988
fbshipit-source-id: a61181f4d6be04c70122b7f79d426e364df1e988
Summary:Source maps are broken on Genymotion right now as they aren't being loaded from the correct URL. refer - https://github.com/facebook/react-native/issues/5338#issuecomment-188232402
**Test plan**
Build and install UIExplorer from master branch in genymotion and enable hot reload. When you change a file and save it, you'll see a Yellow box due to source map fetching failed, as per the referenced comment.
Doing the same for this branch doesn't produce any yellow boxes.
Closes https://github.com/facebook/react-native/pull/6594
Differential Revision: D3088218
Pulled By: martinbigio
fb-gh-sync-id: 0d1c19cc263de5c6c62061c399eef33fa4ac4a7b
shipit-source-id: 0d1c19cc263de5c6c62061c399eef33fa4ac4a7b
Summary:npm-publish.js is not cohesive enough: besides building and publishing it also modifies some files.
It is better to have a separate script that will bump versions, make a commit and tag it.
scripts/bump-oss-version.js does exactly that.
This simplifies release process and allows manual release to npm if CI is not available.
Closes https://github.com/facebook/react-native/pull/6625
Differential Revision: D3092849
fb-gh-sync-id: 92cf38bd3df31c8c9c433fc5f9e15c129855fe0e
shipit-source-id: 92cf38bd3df31c8c9c433fc5f9e15c129855fe0e
Summary:This is the first from the series of PRs I'm going to be sending shorty that would let Animated.js animations to run off the JS thread (for Android only).
This PR introduce a new native module that will be used for offloading animations - NativeAnimatedModule. It has a simple API that allows for animated nodes management via methods like: create/drop animated node, connect/disconnect nodes, start animation of a value node, attach/detach animated from a native view.
Similarly to how we handle UIManager view hierarchy updates we create a queue of animated graph operations that are then executed on the UI thread. This isolates us from problems that may be caused by concurrent updates of animated graph while UI thread is "executing" the animation.
The most important class NativeAnimatedNodesManager.java implements a management interface for animated nodes graph as well as implements a graph traversal algorithm that is run for each animation frame. For each animation frame we visit animated nodes th
Closes https://github.com/facebook/react-native/pull/6466
Differential Revision: D3092739
Pulled By: astreet
fb-gh-sync-id: 665b49900b7367c91a93b9d8864f78fb90bb36ba
shipit-source-id: 665b49900b7367c91a93b9d8864f78fb90bb36ba
Summary: The computation for calculateClippingRect didn't work if you had a horizontal scroll nested in a vertical scroll view. The fix is to simplify the logic to first intersect the inner view within its parent and then apply transforms (translation by top/left and scroll offset). That now provides the correct clipping rectangle for nested views.
Reviewed By: astreet
Differential Revision: D3087367
fb-gh-sync-id: efdd430e024c4599189ddb8dbb258fd2b118690f
shipit-source-id: efdd430e024c4599189ddb8dbb258fd2b118690f
Summary: Adds support to generate stack traces that can be used with source maps that have a `x_facebook_offsets` field. If a file name of a stack frame contains only digits and ends with “.js”, it is included in the trace format generated by `ExceptionsManagerModule`
Reviewed By: martinbigio
Differential Revision: D3072953
fb-gh-sync-id: 69a57e3e2c758034939e5264008871f38f48a78f
shipit-source-id: 69a57e3e2c758034939e5264008871f38f48a78f
Summary:Run `Toast.makeText` other than UI Thread will cause bug #5599 : toast doesn't disappear after app exited.
use `UiThreadUtil.runOnUiThread` to fix this.
Closes https://github.com/facebook/react-native/pull/6443
Differential Revision: D3047261
fb-gh-sync-id: 0096879f2a4b4d76bda996a32089f068ae68e3f3
shipit-source-id: 0096879f2a4b4d76bda996a32089f068ae68e3f3
Summary: When I first did Modal on Android, I incorrectly set the styleWidth and styleHeight on the ModalHostShadowNode. This corresponded to RCTModalHostView in Modal.js. This node is actually really just a dummy node in the tree. The proper node to set the width and height on is the inner <View/> that has top and left position set on it. This updates the code to set the width and height on that inner node.
Reviewed By: mkonicek
Differential Revision: D3077415
fb-gh-sync-id: e9aee0a21333ed0b5bdde11f453381b0a13470c9
shipit-source-id: e9aee0a21333ed0b5bdde11f453381b0a13470c9
Summary:It didn't work for a few reason. First, the drawer view NEEDS to have a background color or no shadow will ever render. Second, we need to use the `setDrawerElevation` method instead of `setElevation` for DrawerLayout. Finally we need to actually pass the style value (maybe we could just pass elevation but I don't really think it can cause any issues) down to the native component as it is not the case at the moment.
I also added a default style to elevation of 16 which is the standard for material design according to https://www.google.com/design/spec/patterns/navigation-drawer.html#navigation-drawer-specs. I could also default it to 0 so it keeps the same appearance as before but I think it looks better this way.
Closes#6022
**Test plan**
Tested using the DrawerLayout in the UIExplorer app.
Before, elevation 0
<img width="420" alt="screen shot 2016-02-23 at 1 55 42 am" src="https://cloud.githubusercontent.com/assets/2677334/13244000/008afdb2-d9d1-11e5-95b8-9c345ea0ea8d.png">
After, elevation
Closes https://github.com/facebook/react-native/pull/6100
Reviewed By: bestander
Differential Revision: D3012242
Pulled By: lexs
fb-gh-sync-id: 4967d7ec920f0229d823032ba95c8a3cace329c6
shipit-source-id: 4967d7ec920f0229d823032ba95c8a3cace329c6
Summary:Adds a `center` option to `Image`'s `resizeMode` prop, which doesn't enlarge images.
This is how it looks in UIExplorer:
{F60386921}
Reviewed By: dmmiller
Differential Revision: D3064284
fb-gh-sync-id: 79cd2da8f44c5b3da2e42d3bebf3131335f53c28
shipit-source-id: 79cd2da8f44c5b3da2e42d3bebf3131335f53c28
Summary: Folly deprecated the brace initialization for dynamic::array, so point to a more recent tarball and fix it.
Reviewed By: bestander
Differential Revision: D3066402
fb-gh-sync-id: aaf70423886cb0cbc14aba7878a1d4e32bce82b1
shipit-source-id: aaf70423886cb0cbc14aba7878a1d4e32bce82b1
Summary:Changed the implementation of `setTranslucent` because the old one used the view flag SYSTEM_UI_FLAG_LAYOUT_FULLSCREEN which has the sad side effect of making the ScrollViews not adjust properly when the keyboard is opened. Instead I use `setOnApplyWindowInsetsListener` to hook in the insets calculation process and consume the top offset so the decor view doesn't add top padding for the status bar.
I also limited the translucent prop to API 21+ because `setOnApplyWindowInsetsListener` is not available before that and anyway the translucent prop is only useful with a semi-transparent status bar color which only works on API 21+.
**Test plan**
Reproduced the bug using this code in UIExplorer.
https://gist.github.com/janicduplessis/217b9994e35f960a1793
Tapping a text field that would be hidden by the soft keyboard should scroll the view so it is not hidden.
Also tested that setting the translucent prop on and off still works as before.
Fixes#6455
Closes https://github.com/facebook/react-native/pull/6481
Differential Revision: D3067199
Pulled By: mkonicek
fb-gh-sync-id: aa115f8688ac7e461e62c18ebb8ab77350d000f8
shipit-source-id: aa115f8688ac7e461e62c18ebb8ab77350d000f8
Summary:This adds a `HEIGHT` constant on `StatusBar` on Android. I needed only this for now but I will work on a better status bar dimensions API later (see TODO).
It also improves the implementation to fix a bug that happened when multiple `StatusBar` components get updated in the same frame as well as remove useless calls to the `StatusBarModule` when values did not change.
Instead of calling the `StatusBarManager` immediately when the component gets updated and relying on the order of the calls that get dispatched to native we now wait at the end of the frame to send the calls to the `StatusBarManager` using `setImmediate`. To make this work properly we need to change the data structure of the props stack a little bit to store the desired transition/animation too for each value.
Finally this updates the example to only show the ones that work for the current platform.
**Test plan**
In the UIExplorer Example, in the 'StatusBar dimensions' section it should show 25 for the height of the status bar.
A
Closes https://github.com/facebook/react-native/pull/6195
Differential Revision: D3017559
fb-gh-sync-id: d6f4c6a72a2dfde83496ecc0f56dca4abaf3055e
shipit-source-id: d6f4c6a72a2dfde83496ecc0f56dca4abaf3055e
Summary:Just added a pass through to the `WebView` for `mediaPlaybackRequiresUserAction` and `setMediaPlaybackRequiresUserGesture` to allow auto-playing audio and video elements
Closes https://github.com/facebook/react-native/pull/5956
Differential Revision: D3053554
Pulled By: mkonicek
fb-gh-sync-id: a1f362c1551de1a0218f5d23c70668e4c8078993
shipit-source-id: a1f362c1551de1a0218f5d23c70668e4c8078993
Summary:Don't add systems current time to the `timeout` option. This is already correct on iOS.
Closes https://github.com/facebook/react-native/pull/6376
Differential Revision: D3053728
Pulled By: mkonicek
fb-gh-sync-id: 919068659d0d23906d8a6d85df7819296460b719
shipit-source-id: 919068659d0d23906d8a6d85df7819296460b719
Summary: D2922108 / PR #5814 broke the “Unbundle” loader by prepending the file path to the entry file with "file://". This converts it back.
Reviewed By: lexs
Differential Revision: D3052749
fb-gh-sync-id: 0b241c50e25f1a1c85265915bee01582d0ba7ac6
shipit-source-id: 0b241c50e25f1a1c85265915bee01582d0ba7ac6
Summary:This is a follow up of 9b87e6c860.
- Allows custom headers on connection request
- Adds a default `origin` header to Android, just like iOS
**Introduces no breaking changes.**
I was working on something similar and would like to propose a few changes that make the API more consistent across both iOS and Android platforms and brings this closer to [spec](https://tools.ietf.org/html/rfc6455).
I believe aprock first implementation of adding custom `headers` was correct. It makes sense naming this argument `headers` since we have no other general options available, and the current `options` field is being used to pass in a header anyway.
My use case for custom headers was attaching a token to the `Authorization` header on the connection request. I have been testing this by passing a JWT inside the `Authorization` header and verifying it on the server before establishing a connection.
Closes https://github.com/facebook/react-native/pull/6016
Differential Revision: D3040735
fb-gh-sync-id: 183744d2415b895f9d9fd8ecf6023a546e18a546
shipit-source-id: 183744d2415b895f9d9fd8ecf6023a546e18a546
Summary:This is a follow up of 9b87e6c860.
- Allows custom headers on connection request
- Adds a default `origin` header to Android, just like iOS
**Introduces no breaking changes.**
I was working on something similar and would like to propose a few changes that make the API more consistent across both iOS and Android platforms and brings this closer to [spec](https://tools.ietf.org/html/rfc6455).
I believe aprock first implementation of adding custom `headers` was correct. It makes sense naming this argument `headers` since we have no other general options available, and the current `options` field is being used to pass in a header anyway.
My use case for custom headers was attaching a token to the `Authorization` header on the connection request. I have been testing this by passing a JWT inside the `Authorization` header and verifying it on the server before establishing a connection.
Closes https://github.com/facebook/react-native/pull/6016
Differential Revision: D3040735
Pulled By: nicklockwood
fb-gh-sync-id: f81bd14ccbdba36309b9d4b4850fb66fe4deae11
shipit-source-id: f81bd14ccbdba36309b9d4b4850fb66fe4deae11
Summary:Turns out even using nanoseconds as timestamps will not guarantee that events
won't happen at the same time. This fixes sorting for event comparison.
Differential Revision: D3030540
fb-gh-sync-id: 2630c50ea60a792ea07b1bf1c6cd46a6d9859268
shipit-source-id: 2630c50ea60a792ea07b1bf1c6cd46a6d9859268
Summary: ReactAndroid will need to use the deprecated version of folly::dynamic for the time being.
Reviewed By: bestander
Differential Revision: D3025433
fb-gh-sync-id: fee6beb65d5d0c449f1a08a0d3c593b788905001
shipit-source-id: fee6beb65d5d0c449f1a08a0d3c593b788905001
Summary: In React, ReactCompoundView is supposed to be implemented by a View, but there is no ViewGroup counterpart that allows mixing virual nodes and non-virtual ones (Views) in the same parent. This is needed because TouchTargetHelper always considers child Views when looking for touch target before falling back to View/ReactCompoundView. This works incorrectly for e.g. layout-only / flattened nodes. ReactCompoundViewGroup allow intercepting touch event before it is dispatched to its children. In that sense, ReactCompoundView.reactTagForTouch() is like View.onTouchEvent() whereas ReactCompoundViewGroup.interceptsTouchEvent() is like ViewGroup.onInterceptTouchEvent().
Differential Revision: D3018028
fb-gh-sync-id: d2c70a55afb9ce9823275e7483d72e0ebedf52e4
shipit-source-id: d2c70a55afb9ce9823275e7483d72e0ebedf52e4
Summary:- Motivation: The WebSocket implementation on Android crashes the app when an attempt is made to write on a web socket that was closed due to a spotty connection. We found this issue by using Pusher, which is built on WebSockets. The following stack trace reveals that the WebSocketModule doesn't catch the case of a closed connection, when a consumer attempts to write:
```sh
Fatal Exception: java.lang.IllegalStateException: closed
at com.squareup.okhttp.internal.ws.RealWebSocket.sendMessage(RealWebSocket.java:109)
at com.facebook.react.modules.websocket.WebSocketModule.send(WebSocketModule.java:176)
at java.lang.reflect.Method.invoke(Method.java)
at java.lang.reflect.Method.invoke(Method.java:372)
at com.facebook.react.bridge.BaseJavaModule$JavaMethod.invoke(BaseJavaModule.java:249)
at com.facebook.react.bridge.NativeModuleRegistry$ModuleDefinition.call(NativeModuleRegistry.java:158)
at com.facebook.react.bridge.NativeModuleRegistry.call(NativeModuleReg
Closes https://github.com/facebook/react-native/pull/6301
Differential Revision: D3016099
fb-gh-sync-id: 838dd9d2e5e5b7a4e2242fa6de5658dfdaf24f55
shipit-source-id: 838dd9d2e5e5b7a4e2242fa6de5658dfdaf24f55
Summary: Split dispatchViewUpdates into two methods, which enables subclasses to commit pending ui operations, even when no root node is present.
Differential Revision: D3011191
fb-gh-sync-id: a3491179441223aeffe21ff483dda582053768e7
shipit-source-id: a3491179441223aeffe21ff483dda582053768e7
Summary: Calling setTimeout(f, 0) will currently schedule f to be called the next frame instead of immediately (which is how it behaves on iOS). This immediately calls back to JS and invokes the function.
Reviewed By: astreet, tadeuzagallo
Differential Revision: D3006125
fb-gh-sync-id: 9fa109ed82836a718cbb2e8cb21da4943d96f5f6
shipit-source-id: 9fa109ed82836a718cbb2e8cb21da4943d96f5f6
Summary:Making buck rebundle the worker script on every JS change is insanely slow. This allows the script to be downloaded for debug builds.
The plan is to couple this with an implementation of `require.resolve` which will automatically insert the correct packager network path in DEV builds and the correct local path in release builds.
e.g.
var worker = new Worker(require.resolve('WebWorkerSample_getPrimesBetween.js'));
Reviewed By: lexs
Differential Revision: D2939279
fb-gh-sync-id: fbf64bbf1df1649b44e4b98ac504d095c10104a6
shipit-source-id: fbf64bbf1df1649b44e4b98ac504d095c10104a6
Summary:There was a deadlock in the bridge if a native module tried to dispatch an event through EventDispatcher (that thread would hold the mTeardownLock and want the mEventStaging lock) at the same time the EventDispatcher callback was triggered and tried to dispatch a call through JS (that thread would hold the mEventStaging lock and want the mTeardownLock).
Now there are two locks (lol). In the scenario above, the native module would hold the mJSToJavaTeardownLock and want the mEventStaging lock, while the EventDispatcher callback would hold the mEventStaging lock and want the mJavaToJSTeardownLock.
Reviewed By: lexs
Differential Revision: D3011526
fb-gh-sync-id: c3ebd5c14a6370d73caebf6c99fcba18a86c6ac1
shipit-source-id: c3ebd5c14a6370d73caebf6c99fcba18a86c6ac1
Summary: Updates the sample to use native modules.
Differential Revision: D2932319
fb-gh-sync-id: 9dc12789839ec2d65e98205508d958d0c8c5b46e
shipit-source-id: 9dc12789839ec2d65e98205508d958d0c8c5b46e
Summary: Example of a conversion to web worker support using the ExecutionContext API changes made in the last set of web worker diffs. WebWorkerSample now creates timers to show that we can dispatch timer calls to multiple JS contexts.
Reviewed By: lexs
Differential Revision: D2928657
fb-gh-sync-id: 17c5f8cd7c63624da43383da7c4160dc48482fe5
shipit-source-id: 17c5f8cd7c63624da43383da7c4160dc48482fe5
Summary: The bridge could be destructed, meaning we can't try to resolve member variables of PlatformBridgeCallback.
Reviewed By: lexs
Differential Revision: D3005994
fb-gh-sync-id: f83de432148ce62fd1060c6b84ec39ad2fa10a55
shipit-source-id: f83de432148ce62fd1060c6b84ec39ad2fa10a55
Summary: This will allow them to clean up resources when a web worker goes away.
Reviewed By: mhorowitz, lexs
Differential Revision: D2994721
fb-gh-sync-id: c7ca1afc7290e85038cf692a139f6478dba0ef61
shipit-source-id: c7ca1afc7290e85038cf692a139f6478dba0ef61
Summary:Added ability to run instrumentation tests with BUCK.
This change uses BUCK to build and run instrumentation tests facebook style.
The gains are that we can execute the same tests internally at FB and in OSS.
Also running tests not via graddle:connect command is 1.5 minutes faster.
I'll keep keep an eye on stability Gradle and BUCK builds for a while.
Closes https://github.com/facebook/react-native/pull/6176
Differential Revision: D2999878
Pulled By: bestander
fb-gh-sync-id: d715ba231769e57100685a1256f2e530c589921c
shipit-source-id: d715ba231769e57100685a1256f2e530c589921c
Summary:To support native modules in web workers, native modules need to have an notion of the JS executor/thread that called into them in order to respond via Callback or JS module call to the right executor on the right thread.
ExecutorToken is an object that only serves as a token that can be used to identify an executor/message queue thread in the bridge. It doesn't expose any methods. Native modules Callback objects automatically have this ExecutionContext attached -- JSModule calls for modules that support workers will need to supply an appropriate ExecutorToken when retrieving the JSModule implementation from the ReactContext.
Reviewed By: mhorowitz
Differential Revision: D2965458
fb-gh-sync-id: 6e354d4df8536d40b12d02bd055f6d06b4ca595d
shipit-source-id: 6e354d4df8536d40b12d02bd055f6d06b4ca595d
Summary:In testing, I've found that there's no good way to return stack traces to the server for exceptions that happen in dtor's. If the dtor is not marked nothrow(false), the exceptions are uncatchable (and bubble up as a std::abort without exception info) and if annotated properly, the program instead crashes trying to resume the stack, again a std::abort without exception info.
Instead, I created a separate destroy method that can be called (and protected via fbjni) to make the dtor's no longer execute code that may throw. Note that we don't really expect the code that was previously in ~JSCExecutor() to throw, but it was in production and we had absolutely no info to help debug it.
Reviewed By: mhorowitz
Differential Revision: D2989999
fb-gh-sync-id: 4cf9de5e0592fe6830a9903375363a78e1339a94
shipit-source-id: 4cf9de5e0592fe6830a9903375363a78e1339a94
Summary: Expose hasUnseenUpdates in ReactShadowNode. Various text input cleanup and fixes.
Differential Revision: D2975870
fb-gh-sync-id: 26c4c0d36e82aaa025b4760603cc48c47ebf46a3
shipit-source-id: 26c4c0d36e82aaa025b4760603cc48c47ebf46a3
Summary: Change the default handling of openURL to attach the NEW_TASK flag if the package we will be launching is different than the currently running package.
Reviewed By: foghina
Differential Revision: D2977441
fb-gh-sync-id: 01d1ac1d791345f815bfc9e8358bce6420c08c1b
shipit-source-id: 01d1ac1d791345f815bfc9e8358bce6420c08c1b
Summary:The implementation of ARTGroupShadowNode saved the canvas and then drew the child nodes but did not reset the canvas afterwards, unlike the behaviour of the other ART shadow nodes. Because of this the matrix operations were compounded for sibling nodes in the surface.
As an example the following code should draw a green circle in the bottom right corner of the surface and a red circle in the top left, which it does on iOS; on Android you'll find that the red circle is drawn in the bottom right corner instead.
```
'use strict';
import React, {
AppRegistry,
Component,
StyleSheet,
View,
} from 'react-native';
const { Surface, Group, Shape } = React.ART;
const styles = StyleSheet.create({
container: {
flex: 1,
justifyContent: 'center',
alignItems: 'center',
backgroundColor: '#808080',
},
surface: {
backgroundColor: '#ffffff'
},
});
class ARTGroupBug extends Component {
render() {
return (
<View style={styles.container}>
<
Closes https://github.com/facebook/react-native/pull/5619
Differential Revision: D2981865
Pulled By: spicyj
fb-gh-sync-id: 95ca701354fe728ed6ee78f8c02d1e1eb70c181e
shipit-source-id: 95ca701354fe728ed6ee78f8c02d1e1eb70c181e
Summary:WRAP_CONTENT is not a valid size for MeasureSpec. While 0 is
a valid value size to pass to makeMeasureSpec, it's slightly cheaper to just
pass in 0 instead since makeMeasureSpec(0, UNSPECIFIED) is 0 anyway.
Differential Revision: D2968322
fb-gh-sync-id: 2ca0e27b12c1a8263800d96fec32127a4b967497
shipit-source-id: 2ca0e27b12c1a8263800d96fec32127a4b967497
Summary:Previously images that have exif orientation data would be cropped and the exif data
would be lost leading to us displaying pictures rotated in error.
Reviewed By: foghina
Differential Revision: D2971596
fb-gh-sync-id: 40f92e27089455259a7d8b83c92d0cf36367e5df
shipit-source-id: 40f92e27089455259a7d8b83c92d0cf36367e5df
Summary: Instead of dispatching calls to the JS thread in Java, do it in the C++ bridge. This moves us closer to the cxx bridge and will allow us to dispatch to the correct web worker in C++ instead of in Java
Reviewed By: mhorowitz
Differential Revision: D2954115
fb-gh-sync-id: 7e7d4eff2c72601b8b4416f1ccd8d2985aebd755
shipit-source-id: 7e7d4eff2c72601b8b4416f1ccd8d2985aebd755
Summary:build-break
Proguard was renaming MessageQueueThread interface methods that we reference from c++.
Differential Revision: D2960332
fb-gh-sync-id: 572dcd8a64e774f65c5abfb8cdf891efcb2bd591
shipit-source-id: 572dcd8a64e774f65c5abfb8cdf891efcb2bd591
Krzysztof Magiera
Micha? Gregorczyk
Emil Sjolander
Maxi Ferreira
Scott Foggo
Steven Luscher
Andy Street
Ahmed El-Helw
zxcpoiu
Lucas Feliciano
Alexander Blom
Felix Oghina
digeff
Martin Konicek
Andrei Coman
Nick
Adam Miskiewicz
Janic Duplessis
Pieter De Baets
VonD
Dmitry Gladkov
Atticus White
Nathan Spaun
Sokovikov
Martin Bigio
Marc Horowitz
Satyajit Sahoo
Konstantin Raev
Saurabh Aggarwal
Dave Miller
David Aurelio
Spencer Ahrens
Fada Chen
DengYun
Bruno Barbieri
Kevin McCray
Brent Vatne
Petr Glotov
Alon Schwarz
Dan Caspi
Michael Lee
Thomas Beverley
Alexander Andersson
Kevin Stumpf
Alexey Dodonov
Christopher Dro
Zen Xu
Chris Hopman
Aaron Chiu
farwayer
Denis Koroskin
Charles Dick
Corentin Smith
AbilashK
Matthew Denner
Jason Sendros
Andreas Drivenes
Konstantin Raev
Mark Hahnenberg
