Summary:
[public / experimental API breaking change]
The data type of `scene.navigationState` is `NavigationRoute`. Rename `scene.navigationState` to
`scene.route` to avoid confusion such as treating `scene.navigationState` as the actual global navigation
state (type: NavigationState).
Reviewed By: ericvicenti
Differential Revision: D3331076
fbshipit-source-id: 3ed989cc8492d398cbeb1b12186459deb261d1fb
Summary:
The default itemPositioning is `automatic` (referred to `auto` in this pull request) - you can check its behaviour in the docs attached.
Sometimes that value has to be modified to have more predictable appearance as described in #4136.
Closes https://github.com/facebook/react-native/pull/7217
Differential Revision: D3220958
Pulled By: mkonicek
fbshipit-source-id: d4bf648b16e71825cd31c06d6b6396479767d19f
Summary:
This is the first step to clarify and simplify the type definations about navigation state.
For now, `NavigationParentState` is actually used as the real navigation state and `NavigationState` is used
as a route in navigation, which has been confusion among the APIs.
To be clear, our APIs has no intention and interest in dealing with nested or hierarchical navigation states,
and we should avoid have the name like `ParentState` or `children`.
To fully migrate the types, theer will be a lot of code changes and this is just the first step to rename.
= What's Next?
1. rename `navigationState.children` to `navigationState.routes` (breaking change!)
2. remove `navigationState.key` from its type defination.
Reviewed By: ericvicenti
Differential Revision: D3321403
fbshipit-source-id: 3e39b60f736c1135bc85d8bf2b89027d665e28d4
Summary:
See "D3291793 Symbolicate JS stacktrace using RN Packager" for the background
of why this matters. TLDR - saving tons of memory by moving symbolication from
JSC into server and cleaning up old hacks.
Before this change: a redbox causes +73MB JSC size
{F60869250}
{F60869249}
After this change: a redbox causes +1MB JSC size
{F61005061}
{F61005062}
Next step – replace JS implementation by native to show better progress and
clean up even more old APIs (ExceptionsManager.updateExceptionMessage).
Reviewed By: davidaurelio
Differential Revision: D3319151
fbshipit-source-id: 48ff4df27642ea4e1bc2414f48a8dd4d32adee50
Summary: - Removed some styling from `SwipeableRow` that wasn't doing much and made slide out view full height
Reviewed By: fkgozali
Differential Revision: D3322849
fbshipit-source-id: 811eee9032c142c61d303ae7e966d8ef7903adaf
Summary:
Before:
- Android had the slideout row flash upon render due to it being rendered first
- iOS had the left side of each row load first, then rerender to show entire row when `scrollViewWidth` is available
Reason:
- Android was loading the slideout view first without an opacity check
- iOS was loading the swipeable view with width 0 first then stretching to `scrollViewWidth` when it was available via `onLayout`
Fix:
Render swipeable view with `flex: 1` then render slideout view
Reviewed By: fkgozali
Differential Revision: D3321466
fbshipit-source-id: 92a3b5e22034e06d05986ddb8c348796bafbbf34
Summary:
- Replace some fixes that were accidentally lost in local rebase that prevent jumpiness when incremental is disabled.
- Require each row to have a key specified by the caller to prevent jumping because of accidental duplicates or unneeded/problematic row re-rendering because of legit re-ordering.
Reviewed By: steveluscher
Differential Revision: D3322006
fbshipit-source-id: 0019aab07cb1fe2b148a14b5818de53aa373eb50
Summary:
Incremental rendering is a tradeoff between throughput and responsiveness because it yields. When we have plenty of
buffer (say 50% of the target), we render incrementally to keep the app responsive. If we are dangerously low on buffer
(say below 25%) we always disable incremental to try to catch up as fast as possible. In between, we only disable
incremental while actively scrolling since it's unlikely the user will try to press a button while scrolling.
This also optimizes some things then incremental is switching back and forth.
I played around with making the render window itself adaptive, but it seems pretty futile to predict - once the user
decides to scroll quickly in some direction, it's pretty much too late and increasing the render window size won't help
because we're already limited by the render throughput at that point.
Reviewed By: ericvicenti
Differential Revision: D3250916
fbshipit-source-id: 930d418522a3bf3e20083e60f6eb6f891497a2b8
Summary: Default image icon size for Quick Actions could cause slideout view to bleed over the slideable view. All styling has been removed and must now be passed in, thus allowing every caller to control size for best fit.
Reviewed By: furdei
Differential Revision: D3315696
fbshipit-source-id: 8f8b3d2cf7a005d42a18d434d9e0080c64597be0
Summary:
- Fork NavigationAnimatedView to NavigationTransitioner
- NavigationAnimatedView will soon be deprecated and we'd ask people to use NavigationTransitioner instead.
Difference between NavigationTransitioner and NavigationAnimatedView
- prop `applyAnimation` is removed.
- new prop `configureTransition`, `onTransitionStart`, and `onTransitionEnd` are added.
tl;dr;
In NavigationAnimatedView, we `position` (an Animated.Value object) as a proxy of the
transtion which happens whenever the index of navigation state changes.
Because `position` does not change unless navigation index changes, it won't
be possible to build animations for actions that changes the navigation state
without changing the index.
Also, we believe that the name `Transitioner` is a better name for this core component
that focuses on transitioning. Note that the actual animation work is done via
`<Animated.View />` returnd from the `renderScene` prop.
Reviewed By: ericvicenti
Differential Revision: D3302688
fbshipit-source-id: 720c3a4d3ccf97eb05b038baa44c9e780aad120b
Summary:
... used as both a shape and plain object.
this splits them out so both parts can be used as needed.
NavigationPropTypes.SceneRenderer is a PropTypes shape
NavigationPropTypes.SceneRendererProps is the plain object that makes up the shape.
Closes https://github.com/facebook/react-native/pull/7518
Differential Revision: D3317322
Pulled By: ericvicenti
fbshipit-source-id: e8a31e05130e6647b63f68dbef31bc874550948c
Summary:
This works with D3305161 to minimize image flashing. After D3305161, the completion handler passed to `-[RCTImageLoader loadImageWithoutClipping:size:scale:resizeMode:progressBlock:completionBlock:]` may be called back on the main queue in the case of a cached image. In this case, we want to set the image view's image property synchronously rather than on the next runloop iteration via dispatch_async. This minimizes the amount of image flashing the user sees when displaying a cached image.
The exception to this case is for blurred images. A blur can be an expensive (taking multiple ms on the CPU), so we always make sure to perform the blur off the main queue even if the image is cached and the callback came back on the main queue.
Reviewed By: nicklockwood
Differential Revision: D3310176
fbshipit-source-id: 6820782527b65e4956879cf06e8ed2c09c622a58
Summary:
Previously, only Text and Image could be nested within Text. Now, any
view can be nested within Text. One restriction of this feature is
that developers must give inline views a width and a height via
the style prop.
Previously, inline Images were supported by using iOS's built-in support
for rendering images with an NSAttributedString via NSTextAttachment.
However, NSAttributedString doesn't support rendering arbitrary views.
This change adds support for nesting views within Text by creating one
NSTextAttachment per inline view. The NSTextAttachments act as placeholders.
They are set to be the size of the corresponding view. After the text is
laid out, we query the text system to find out where it has positioned each
NSTextAttachment. We then position the views to be at those locations.
This commit also contains a change in `RCTShadowText.m`
`_setParagraphStyleOnAttributedString:heightOfTallestSubview:`. It now only sets
`lineHeight`, `textAlign`, and `writingDirection` when they've actua
Closes https://github.com/facebook/react-native/pull/7304
Differential Revision: D3269333
Pulled By: nicklockwood
fbshipit-source-id: 2b59f1c5445a4012f9c29df9f10f5010060ea517
Summary:
This fixes an issue where animations for values near zero could end up formatted
with exponents (e.g. `1.452e-10`), which is not valid for an `rgba` color spec.
This commit arbitrarily rounds it to the nearest thousandth to prevent this type
of formatting while still maintaining high-enough resolution in the alpha channel.
One way this could bubble up to the user is as PropType validation failures:
```
Failed propType: Invalid prop `backgroundColor` supplied to `RCTView`: rgba(0, 0, 0, 9.838983123336224e-7)
```
Closes https://github.com/facebook/react-native/pull/7597
Differential Revision: D3310941
Pulled By: vjeux
fbshipit-source-id: 0c95facaef5b69c021662af9fb6f268d890ecc3e
Summary:
- Removed unnecessary rerending of `SwipeableListView` by properly managing `SwipeableListViewDataSource`
- Simplified `SwipeableRow` logic and improved swiping performance
- Added bounce effect
- Locked `ListView` from being scrollable when `SwipeableRow` is being swiped; behaviour mirrors that of Android on iOS and significantly improves framerates
Reviewed By: fkgozali
Differential Revision: D3307599
fbshipit-source-id: 168b6b72ef1f9e47d0145cf9e1baecbab3564b84
Summary:
Sometimes list view row components(even the ones unseen) have a frame but its x and y is 0 or uninitialized at view loading time, which will lead to _updateVisibleRows firing for all the row components incorrectly at loading time.
This is to make the code more defensive.
Reviewed By: sahrens
Differential Revision: D3308515
fbshipit-source-id: 0e30fdf10345629bfd9de3fe0ad641bb1903fe00
Summary:
The function should have optional arguments, but not nullable types.
Closes https://github.com/facebook/react-native/pull/7580
Differential Revision: D3304994
Pulled By: davidaurelio
fbshipit-source-id: 9918817429126ea16dd3076b61a3493c3a3a8388
Summary:
The `EmitterSubscription.remove()` method was previously calling `this.subscriber.removeSubscription(this)` directly, bypassing the mechanism in `NativeEventEmitter` that keeps track of the number of subscriptions.
This meant that native event modules (subclasses of `RCTEventEmitter`) would keep sending events even after all the listeners had been removed. This wasn't a huge overhead, since these modules are singletons and only send one message over the bridge per event, regardless of the number of listeners, but it's still undesirable.
This fixes the problem by routing the `EmitterSubscription.remove()` method through the `EventEmitter` so that `NativeEventEmitter` can apply the additional native calls.
I've also improved the architecture so that each `NativeEventEmitter` uses its own `EventEmitter`, but they currently all still share the same `EventSubscriptionVendor` so that legacy code which registers events via `RCTDeviceEventEmitter` still works.
Reviewed By: vjeux
Differential Revision: D3292361
fbshipit-source-id: d60e881d50351523d2112473703bea826641cdef
Summary:
Further describe the methods available on NetInfo.
Question: How do I update the docs on the website?
Closes https://github.com/facebook/react-native/pull/7375
Differential Revision: D3303300
fbshipit-source-id: 4343d490f65e4e47b93f2c98a645cb675d2cf708
Summary:
For now, we emit `didFocus` event before fully resetting the transition and
hiding the stale scenes, so handler for `didFocus` event does not get the right
scenes in place. We should emit the didFocus event after resetting the transition and
hiding the stale scenes.
Reviewed By: ericvicenti
Differential Revision: D3299002
fbshipit-source-id: 6a79528097aabc4f61d4dba322820c228c889a84
Summary:
We want to give people the ability to log scroll performance (including Fb).
This adds an interface that can be enabled and disabled from the react scroll views.
This is a prerequisite to implementing the actual framerate logger that will log dropped
frames while scrolling in prod.
Reviewed By: astreet
Differential Revision: D3283588
fbshipit-source-id: ed9736cb9ed3f441511647f36b1460092bd91e56
Summary:
= Breaking Change (for experimental features) =
Major API changes in NavigationAnimatedView
= New prop `transition` for scene renderer
In NavigationAnimatedView, we should not use `position` as a proxy of the
transtion which happens whenever navigation state changes.
Because `position` does not change unless navigation index changes, it won't
be possible to build animations for actions that replace navigation state
without changing the index.
This diff introduces an abstract prop `transition` that is exposed to the scene
renderers.
= Replace `applyAnimation` with `configureTransition`.
Expose a new optional prop `configureTransition` that allows people to configure
transitions easily.
For instance, to configure the transition, do this:
```
function configureTransition() {
return {
dutation: 123,
easing: Easing.easeInOut,
};
}
```
<NavigationAnimatedView configureTransition={configureTransition) />
```
Reviewed By: ericvicenti
Differential Revision: D3278698
fbshipit-source-id: b790b92e0fabb42488ff1135b1c37a3f0e9420f7
Summary: Center align text on Android
Reviewed By: fkgozali
Differential Revision: D3295222
fbshipit-source-id: c707d1102c8aa86aa04dc9b34a4746b1ff1543bc
Summary:
= Breaking Change (for experimental features) =
Major API changes in NavigationAnimatedView
= New prop `transition` for scene renderer
In NavigationAnimatedView, we should not use `position` as a proxy of the
transtion which happens whenever navigation state changes.
Because `position` does not change unless navigation index changes, it won't
be possible to build animations for actions that replace navigation state
without changing the index.
This diff introduces an abstract prop `transition` that is exposed to the scene
renderers.
= Replace `applyAnimation` with `configureTransition`.
Expose a new optional prop `configureTransition` that allows people to configure
transitions easily.
For instance, to configure the transition, do this:
```
function configureTransition() {
return {
dutation: 123,
easing: Easing.easeInOut,
};
}
```
<NavigationAnimatedView configureTransition={configureTransition) />
```
Reviewed By: ericvicenti
Differential Revision: D3278698
fbshipit-source-id: 25ebad286d8b41f46c35c0f32d6023ebd01f19e7
Summary:
- Implemented a quick actions template to allow children buttons to be rendered in the slideout of `SwipeableListView`
- Implemented `QuickActionsButton` to allow buttons to be used in the slideout view
Reviewed By: fkgozali
Differential Revision: D3271946
fbshipit-source-id: b7c7412f44be64dea7084ca36c4b0f4f3842b008
Summary:
This is a nice feature to have.
I've tested this by copying and renaming the ViewPager java and javascript files from the react-native repo and including them in a project. Whats the best way to test this directly from the repo?
Closes https://github.com/facebook/react-native/pull/5968
Differential Revision: D3240651
Pulled By: mkonicek
fbshipit-source-id: 5f1d157216df4f3314915496188a92aec1b85e91
Summary:
= Breaking Change (for experimental features) =
<NavigationView /> serves no obvious benefit and it's costly to maintain it.
- its features could be completely replaced by NavigationAnimatedView without
enabling the transtion animation.
- it could also be replaced with a simple view and normal scene renderer.
Reviewed By: ericvicenti
Differential Revision: D3280547
fbshipit-source-id: 4ac3dbe92ceb9107a1f6e77a78bd6021485e78a9
Summary:
This is a solution for the problem I raised in https://www.facebook.com/groups/react.native.community/permalink/768218933313687/
I've added a new native base class, `RCTEventEmitter` as well as an equivalent JS class/module `NativeEventEmitter` (RCTEventEmitter.js and EventEmitter.js were taken already).
Instead of arbitrary modules sending events via `bridge.eventDispatcher`, the idea is that any module that sends events should now subclass `RCTEventEmitter`, and provide an equivalent JS module that subclasses `NativeEventEmitter`.
JS code that wants to observe the events should now observe it via the specific JS module rather than via `RCTDeviceEventEmitter` directly. e.g. to observer a keyboard event, instead of writing:
const RCTDeviceEventEmitter = require('RCTDeviceEventEmitter');
RCTDeviceEventEmitter.addListener('keyboardWillShow', (event) => { ... });
You'd now write:
const Keyboard = require('Keyboard');
Keyboard.addListener('keyboardWillShow', (event) => { ... });
Within a component, you can also use the `Subscribable.Mixin` as you would previously, but instead of:
this.addListenerOn(RCTDeviceEventEmitter, 'keyboardWillShow', ...);
Write:
this.addListenerOn(Keyboard, 'keyboardWillShow', ...);
This approach allows the native `RCTKeyboardObserver` module to be created lazily the first time a listener is added, and to stop sending events when the last listener is removed. It also allows us to validate that the event strings being observed and omitted match the supported events for that module.
As a proof-of-concept, I've converted the `RCTStatusBarManager` and `RCTKeyboardObserver` modules to use the new system. I'll convert the rest in a follow up diff.
For now, the new `NativeEventEmitter` JS module wraps the `RCTDeviceEventEmitter` JS module, and just uses the native `RCTEventEmitter` module for bookkeeping. This allows for full backwards compatibility (code that is observing the event via `RCTDeviceEventEmitter` instead of the specific module will still work as expected, albeit with a warning). Once all legacy calls have been removed, this could be refactored to something more elegant internally, whilst maintaining the same public interface.
Note: currently, all device events still share a single global namespace, since they're really all registered on the same emitter instance internally. We should move away from that as soon as possible because it's not intuitive and will likely lead to strange bugs if people add generic events such as "onChange" or "onError" to their modules (which is common practice for components, where it's not a problem).
Reviewed By: javache
Differential Revision: D3269966
fbshipit-source-id: 1412daba850cd373020e1086673ba38ef9193050
Corné Dorrestijn
Hedger Wang
Mike Grabowski
Alex Kotliarskyi
dmueller39
Fred Liu
Spencer Ahrens
Alexander Blom
Joe Noon
Alex Kotliarskyi
Ben Nham
Adam Comella
Sean Kelley
Yu Huang
Tore Hanssen
Naman Goel
Nick Lockwood
Yao Bin Then
Jimmy Mayoukou
Nathan Spaun
Olivier Notteghem
Pieter De Baets
Kevin Johnson
