mirror of
https://github.com/zebrajr/react.git
synced 2025-12-07 12:20:38 +01:00
* Add Event Replaying Infra * Wire up Roots and Suspense boundaries, to retry events, after they commit * Replay discrete events in order in a separate scheduler callback * Add continuous events These events only replay their last target if the target is not yet hydrated. That way we don't have to wait for a previously hovered boundary before invoking the current target. * Enable tests from before These tests were written with replaying in mind and now we can properly enable them. * Unify replaying and dispatching * Mark system flags as a replay and pass to legacy events That way we can check if this is a replay and therefore needs a special case. One such special case is "mouseover" where we check the relatedTarget. * Eagerly listen to all replayable events To minimize breakages in a minor, I only do this for the new root APIs since replaying only matters there anyway. Only if hydrating. For Flare, I have to attach all active listeners since the current system has one DOM listener for each. In a follow up I plan on optimizing that by only attaching one if there's at least one active listener which would allow us to start with only passive and then upgrade. * Desperate attempt to save bytese * Add test for mouseover replaying We need to check if the "relatedTarget" is mounted due to how the old event system dispatches from the "out" event. * Fix for nested boundaries and suspense in root container This is a follow up to #16673 which didn't have a test because it wasn't observable yet. This shows that it had a bug. * Rename RESPONDER_EVENT_SYSTEM to PLUGIN_EVENT_SYSTEM
611 lines
24 KiB
JavaScript
611 lines
24 KiB
JavaScript
/**
|
|
* Copyright (c) Facebook, Inc. and its affiliates.
|
|
*
|
|
* This source code is licensed under the MIT license found in the
|
|
* LICENSE file in the root directory of this source tree.
|
|
*/
|
|
|
|
import {getLowestCommonAncestor, isAncestor} from 'shared/ReactTreeTraversal';
|
|
|
|
import {
|
|
executeDirectDispatch,
|
|
hasDispatches,
|
|
executeDispatchesInOrderStopAtTrue,
|
|
getInstanceFromNode,
|
|
} from './EventPluginUtils';
|
|
import {
|
|
accumulateDirectDispatches,
|
|
accumulateTwoPhaseDispatches,
|
|
accumulateTwoPhaseDispatchesSkipTarget,
|
|
} from './EventPropagators';
|
|
import ResponderSyntheticEvent from './ResponderSyntheticEvent';
|
|
import ResponderTouchHistoryStore from './ResponderTouchHistoryStore';
|
|
import accumulate from './accumulate';
|
|
import {
|
|
TOP_SCROLL,
|
|
TOP_SELECTION_CHANGE,
|
|
TOP_TOUCH_CANCEL,
|
|
isStartish,
|
|
isMoveish,
|
|
isEndish,
|
|
startDependencies,
|
|
moveDependencies,
|
|
endDependencies,
|
|
} from './ResponderTopLevelEventTypes';
|
|
|
|
/**
|
|
* Instance of element that should respond to touch/move types of interactions,
|
|
* as indicated explicitly by relevant callbacks.
|
|
*/
|
|
let responderInst = null;
|
|
|
|
/**
|
|
* Count of current touches. A textInput should become responder iff the
|
|
* selection changes while there is a touch on the screen.
|
|
*/
|
|
let trackedTouchCount = 0;
|
|
|
|
const changeResponder = function(nextResponderInst, blockHostResponder) {
|
|
const oldResponderInst = responderInst;
|
|
responderInst = nextResponderInst;
|
|
if (ResponderEventPlugin.GlobalResponderHandler !== null) {
|
|
ResponderEventPlugin.GlobalResponderHandler.onChange(
|
|
oldResponderInst,
|
|
nextResponderInst,
|
|
blockHostResponder,
|
|
);
|
|
}
|
|
};
|
|
|
|
const eventTypes = {
|
|
/**
|
|
* On a `touchStart`/`mouseDown`, is it desired that this element become the
|
|
* responder?
|
|
*/
|
|
startShouldSetResponder: {
|
|
phasedRegistrationNames: {
|
|
bubbled: 'onStartShouldSetResponder',
|
|
captured: 'onStartShouldSetResponderCapture',
|
|
},
|
|
dependencies: startDependencies,
|
|
},
|
|
|
|
/**
|
|
* On a `scroll`, is it desired that this element become the responder? This
|
|
* is usually not needed, but should be used to retroactively infer that a
|
|
* `touchStart` had occurred during momentum scroll. During a momentum scroll,
|
|
* a touch start will be immediately followed by a scroll event if the view is
|
|
* currently scrolling.
|
|
*
|
|
* TODO: This shouldn't bubble.
|
|
*/
|
|
scrollShouldSetResponder: {
|
|
phasedRegistrationNames: {
|
|
bubbled: 'onScrollShouldSetResponder',
|
|
captured: 'onScrollShouldSetResponderCapture',
|
|
},
|
|
dependencies: [TOP_SCROLL],
|
|
},
|
|
|
|
/**
|
|
* On text selection change, should this element become the responder? This
|
|
* is needed for text inputs or other views with native selection, so the
|
|
* JS view can claim the responder.
|
|
*
|
|
* TODO: This shouldn't bubble.
|
|
*/
|
|
selectionChangeShouldSetResponder: {
|
|
phasedRegistrationNames: {
|
|
bubbled: 'onSelectionChangeShouldSetResponder',
|
|
captured: 'onSelectionChangeShouldSetResponderCapture',
|
|
},
|
|
dependencies: [TOP_SELECTION_CHANGE],
|
|
},
|
|
|
|
/**
|
|
* On a `touchMove`/`mouseMove`, is it desired that this element become the
|
|
* responder?
|
|
*/
|
|
moveShouldSetResponder: {
|
|
phasedRegistrationNames: {
|
|
bubbled: 'onMoveShouldSetResponder',
|
|
captured: 'onMoveShouldSetResponderCapture',
|
|
},
|
|
dependencies: moveDependencies,
|
|
},
|
|
|
|
/**
|
|
* Direct responder events dispatched directly to responder. Do not bubble.
|
|
*/
|
|
responderStart: {
|
|
registrationName: 'onResponderStart',
|
|
dependencies: startDependencies,
|
|
},
|
|
responderMove: {
|
|
registrationName: 'onResponderMove',
|
|
dependencies: moveDependencies,
|
|
},
|
|
responderEnd: {
|
|
registrationName: 'onResponderEnd',
|
|
dependencies: endDependencies,
|
|
},
|
|
responderRelease: {
|
|
registrationName: 'onResponderRelease',
|
|
dependencies: endDependencies,
|
|
},
|
|
responderTerminationRequest: {
|
|
registrationName: 'onResponderTerminationRequest',
|
|
dependencies: [],
|
|
},
|
|
responderGrant: {
|
|
registrationName: 'onResponderGrant',
|
|
dependencies: [],
|
|
},
|
|
responderReject: {
|
|
registrationName: 'onResponderReject',
|
|
dependencies: [],
|
|
},
|
|
responderTerminate: {
|
|
registrationName: 'onResponderTerminate',
|
|
dependencies: [],
|
|
},
|
|
};
|
|
|
|
/**
|
|
*
|
|
* Responder System:
|
|
* ----------------
|
|
*
|
|
* - A global, solitary "interaction lock" on a view.
|
|
* - If a node becomes the responder, it should convey visual feedback
|
|
* immediately to indicate so, either by highlighting or moving accordingly.
|
|
* - To be the responder means, that touches are exclusively important to that
|
|
* responder view, and no other view.
|
|
* - While touches are still occurring, the responder lock can be transferred to
|
|
* a new view, but only to increasingly "higher" views (meaning ancestors of
|
|
* the current responder).
|
|
*
|
|
* Responder being granted:
|
|
* ------------------------
|
|
*
|
|
* - Touch starts, moves, and scrolls can cause an ID to become the responder.
|
|
* - We capture/bubble `startShouldSetResponder`/`moveShouldSetResponder` to
|
|
* the "appropriate place".
|
|
* - If nothing is currently the responder, the "appropriate place" is the
|
|
* initiating event's `targetID`.
|
|
* - If something *is* already the responder, the "appropriate place" is the
|
|
* first common ancestor of the event target and the current `responderInst`.
|
|
* - Some negotiation happens: See the timing diagram below.
|
|
* - Scrolled views automatically become responder. The reasoning is that a
|
|
* platform scroll view that isn't built on top of the responder system has
|
|
* began scrolling, and the active responder must now be notified that the
|
|
* interaction is no longer locked to it - the system has taken over.
|
|
*
|
|
* - Responder being released:
|
|
* As soon as no more touches that *started* inside of descendants of the
|
|
* *current* responderInst, an `onResponderRelease` event is dispatched to the
|
|
* current responder, and the responder lock is released.
|
|
*
|
|
* TODO:
|
|
* - on "end", a callback hook for `onResponderEndShouldRemainResponder` that
|
|
* determines if the responder lock should remain.
|
|
* - If a view shouldn't "remain" the responder, any active touches should by
|
|
* default be considered "dead" and do not influence future negotiations or
|
|
* bubble paths. It should be as if those touches do not exist.
|
|
* -- For multitouch: Usually a translate-z will choose to "remain" responder
|
|
* after one out of many touches ended. For translate-y, usually the view
|
|
* doesn't wish to "remain" responder after one of many touches end.
|
|
* - Consider building this on top of a `stopPropagation` model similar to
|
|
* `W3C` events.
|
|
* - Ensure that `onResponderTerminate` is called on touch cancels, whether or
|
|
* not `onResponderTerminationRequest` returns `true` or `false`.
|
|
*
|
|
*/
|
|
|
|
/* Negotiation Performed
|
|
+-----------------------+
|
|
/ \
|
|
Process low level events to + Current Responder + wantsResponderID
|
|
determine who to perform negot-| (if any exists at all) |
|
|
iation/transition | Otherwise just pass through|
|
|
-------------------------------+----------------------------+------------------+
|
|
Bubble to find first ID | |
|
|
to return true:wantsResponderID| |
|
|
| |
|
|
+-------------+ | |
|
|
| onTouchStart| | |
|
|
+------+------+ none | |
|
|
| return| |
|
|
+-----------v-------------+true| +------------------------+ |
|
|
|onStartShouldSetResponder|----->|onResponderStart (cur) |<-----------+
|
|
+-----------+-------------+ | +------------------------+ | |
|
|
| | | +--------+-------+
|
|
| returned true for| false:REJECT +-------->|onResponderReject
|
|
| wantsResponderID | | | +----------------+
|
|
| (now attempt | +------------------+-----+ |
|
|
| handoff) | | onResponder | |
|
|
+------------------->| TerminationRequest| |
|
|
| +------------------+-----+ |
|
|
| | | +----------------+
|
|
| true:GRANT +-------->|onResponderGrant|
|
|
| | +--------+-------+
|
|
| +------------------------+ | |
|
|
| | onResponderTerminate |<-----------+
|
|
| +------------------+-----+ |
|
|
| | | +----------------+
|
|
| +-------->|onResponderStart|
|
|
| | +----------------+
|
|
Bubble to find first ID | |
|
|
to return true:wantsResponderID| |
|
|
| |
|
|
+-------------+ | |
|
|
| onTouchMove | | |
|
|
+------+------+ none | |
|
|
| return| |
|
|
+-----------v-------------+true| +------------------------+ |
|
|
|onMoveShouldSetResponder |----->|onResponderMove (cur) |<-----------+
|
|
+-----------+-------------+ | +------------------------+ | |
|
|
| | | +--------+-------+
|
|
| returned true for| false:REJECT +-------->|onResponderRejec|
|
|
| wantsResponderID | | | +----------------+
|
|
| (now attempt | +------------------+-----+ |
|
|
| handoff) | | onResponder | |
|
|
+------------------->| TerminationRequest| |
|
|
| +------------------+-----+ |
|
|
| | | +----------------+
|
|
| true:GRANT +-------->|onResponderGrant|
|
|
| | +--------+-------+
|
|
| +------------------------+ | |
|
|
| | onResponderTerminate |<-----------+
|
|
| +------------------+-----+ |
|
|
| | | +----------------+
|
|
| +-------->|onResponderMove |
|
|
| | +----------------+
|
|
| |
|
|
| |
|
|
Some active touch started| |
|
|
inside current responder | +------------------------+ |
|
|
+------------------------->| onResponderEnd | |
|
|
| | +------------------------+ |
|
|
+---+---------+ | |
|
|
| onTouchEnd | | |
|
|
+---+---------+ | |
|
|
| | +------------------------+ |
|
|
+------------------------->| onResponderEnd | |
|
|
No active touches started| +-----------+------------+ |
|
|
inside current responder | | |
|
|
| v |
|
|
| +------------------------+ |
|
|
| | onResponderRelease | |
|
|
| +------------------------+ |
|
|
| |
|
|
+ + */
|
|
|
|
/**
|
|
* A note about event ordering in the `EventPluginHub`.
|
|
*
|
|
* Suppose plugins are injected in the following order:
|
|
*
|
|
* `[R, S, C]`
|
|
*
|
|
* To help illustrate the example, assume `S` is `SimpleEventPlugin` (for
|
|
* `onClick` etc) and `R` is `ResponderEventPlugin`.
|
|
*
|
|
* "Deferred-Dispatched Events":
|
|
*
|
|
* - The current event plugin system will traverse the list of injected plugins,
|
|
* in order, and extract events by collecting the plugin's return value of
|
|
* `extractEvents()`.
|
|
* - These events that are returned from `extractEvents` are "deferred
|
|
* dispatched events".
|
|
* - When returned from `extractEvents`, deferred-dispatched events contain an
|
|
* "accumulation" of deferred dispatches.
|
|
* - These deferred dispatches are accumulated/collected before they are
|
|
* returned, but processed at a later time by the `EventPluginHub` (hence the
|
|
* name deferred).
|
|
*
|
|
* In the process of returning their deferred-dispatched events, event plugins
|
|
* themselves can dispatch events on-demand without returning them from
|
|
* `extractEvents`. Plugins might want to do this, so that they can use event
|
|
* dispatching as a tool that helps them decide which events should be extracted
|
|
* in the first place.
|
|
*
|
|
* "On-Demand-Dispatched Events":
|
|
*
|
|
* - On-demand-dispatched events are not returned from `extractEvents`.
|
|
* - On-demand-dispatched events are dispatched during the process of returning
|
|
* the deferred-dispatched events.
|
|
* - They should not have side effects.
|
|
* - They should be avoided, and/or eventually be replaced with another
|
|
* abstraction that allows event plugins to perform multiple "rounds" of event
|
|
* extraction.
|
|
*
|
|
* Therefore, the sequence of event dispatches becomes:
|
|
*
|
|
* - `R`s on-demand events (if any) (dispatched by `R` on-demand)
|
|
* - `S`s on-demand events (if any) (dispatched by `S` on-demand)
|
|
* - `C`s on-demand events (if any) (dispatched by `C` on-demand)
|
|
* - `R`s extracted events (if any) (dispatched by `EventPluginHub`)
|
|
* - `S`s extracted events (if any) (dispatched by `EventPluginHub`)
|
|
* - `C`s extracted events (if any) (dispatched by `EventPluginHub`)
|
|
*
|
|
* In the case of `ResponderEventPlugin`: If the `startShouldSetResponder`
|
|
* on-demand dispatch returns `true` (and some other details are satisfied) the
|
|
* `onResponderGrant` deferred dispatched event is returned from
|
|
* `extractEvents`. The sequence of dispatch executions in this case
|
|
* will appear as follows:
|
|
*
|
|
* - `startShouldSetResponder` (`ResponderEventPlugin` dispatches on-demand)
|
|
* - `touchStartCapture` (`EventPluginHub` dispatches as usual)
|
|
* - `touchStart` (`EventPluginHub` dispatches as usual)
|
|
* - `responderGrant/Reject` (`EventPluginHub` dispatches as usual)
|
|
*/
|
|
|
|
function setResponderAndExtractTransfer(
|
|
topLevelType,
|
|
targetInst,
|
|
nativeEvent,
|
|
nativeEventTarget,
|
|
) {
|
|
const shouldSetEventType = isStartish(topLevelType)
|
|
? eventTypes.startShouldSetResponder
|
|
: isMoveish(topLevelType)
|
|
? eventTypes.moveShouldSetResponder
|
|
: topLevelType === TOP_SELECTION_CHANGE
|
|
? eventTypes.selectionChangeShouldSetResponder
|
|
: eventTypes.scrollShouldSetResponder;
|
|
|
|
// TODO: stop one short of the current responder.
|
|
const bubbleShouldSetFrom = !responderInst
|
|
? targetInst
|
|
: getLowestCommonAncestor(responderInst, targetInst);
|
|
|
|
// When capturing/bubbling the "shouldSet" event, we want to skip the target
|
|
// (deepest ID) if it happens to be the current responder. The reasoning:
|
|
// It's strange to get an `onMoveShouldSetResponder` when you're *already*
|
|
// the responder.
|
|
const skipOverBubbleShouldSetFrom = bubbleShouldSetFrom === responderInst;
|
|
const shouldSetEvent = ResponderSyntheticEvent.getPooled(
|
|
shouldSetEventType,
|
|
bubbleShouldSetFrom,
|
|
nativeEvent,
|
|
nativeEventTarget,
|
|
);
|
|
shouldSetEvent.touchHistory = ResponderTouchHistoryStore.touchHistory;
|
|
if (skipOverBubbleShouldSetFrom) {
|
|
accumulateTwoPhaseDispatchesSkipTarget(shouldSetEvent);
|
|
} else {
|
|
accumulateTwoPhaseDispatches(shouldSetEvent);
|
|
}
|
|
const wantsResponderInst = executeDispatchesInOrderStopAtTrue(shouldSetEvent);
|
|
if (!shouldSetEvent.isPersistent()) {
|
|
shouldSetEvent.constructor.release(shouldSetEvent);
|
|
}
|
|
|
|
if (!wantsResponderInst || wantsResponderInst === responderInst) {
|
|
return null;
|
|
}
|
|
let extracted;
|
|
const grantEvent = ResponderSyntheticEvent.getPooled(
|
|
eventTypes.responderGrant,
|
|
wantsResponderInst,
|
|
nativeEvent,
|
|
nativeEventTarget,
|
|
);
|
|
grantEvent.touchHistory = ResponderTouchHistoryStore.touchHistory;
|
|
|
|
accumulateDirectDispatches(grantEvent);
|
|
const blockHostResponder = executeDirectDispatch(grantEvent) === true;
|
|
if (responderInst) {
|
|
const terminationRequestEvent = ResponderSyntheticEvent.getPooled(
|
|
eventTypes.responderTerminationRequest,
|
|
responderInst,
|
|
nativeEvent,
|
|
nativeEventTarget,
|
|
);
|
|
terminationRequestEvent.touchHistory =
|
|
ResponderTouchHistoryStore.touchHistory;
|
|
accumulateDirectDispatches(terminationRequestEvent);
|
|
const shouldSwitch =
|
|
!hasDispatches(terminationRequestEvent) ||
|
|
executeDirectDispatch(terminationRequestEvent);
|
|
if (!terminationRequestEvent.isPersistent()) {
|
|
terminationRequestEvent.constructor.release(terminationRequestEvent);
|
|
}
|
|
|
|
if (shouldSwitch) {
|
|
const terminateEvent = ResponderSyntheticEvent.getPooled(
|
|
eventTypes.responderTerminate,
|
|
responderInst,
|
|
nativeEvent,
|
|
nativeEventTarget,
|
|
);
|
|
terminateEvent.touchHistory = ResponderTouchHistoryStore.touchHistory;
|
|
accumulateDirectDispatches(terminateEvent);
|
|
extracted = accumulate(extracted, [grantEvent, terminateEvent]);
|
|
changeResponder(wantsResponderInst, blockHostResponder);
|
|
} else {
|
|
const rejectEvent = ResponderSyntheticEvent.getPooled(
|
|
eventTypes.responderReject,
|
|
wantsResponderInst,
|
|
nativeEvent,
|
|
nativeEventTarget,
|
|
);
|
|
rejectEvent.touchHistory = ResponderTouchHistoryStore.touchHistory;
|
|
accumulateDirectDispatches(rejectEvent);
|
|
extracted = accumulate(extracted, rejectEvent);
|
|
}
|
|
} else {
|
|
extracted = accumulate(extracted, grantEvent);
|
|
changeResponder(wantsResponderInst, blockHostResponder);
|
|
}
|
|
return extracted;
|
|
}
|
|
|
|
/**
|
|
* A transfer is a negotiation between a currently set responder and the next
|
|
* element to claim responder status. Any start event could trigger a transfer
|
|
* of responderInst. Any move event could trigger a transfer.
|
|
*
|
|
* @param {string} topLevelType Record from `BrowserEventConstants`.
|
|
* @return {boolean} True if a transfer of responder could possibly occur.
|
|
*/
|
|
function canTriggerTransfer(topLevelType, topLevelInst, nativeEvent) {
|
|
return (
|
|
topLevelInst &&
|
|
// responderIgnoreScroll: We are trying to migrate away from specifically
|
|
// tracking native scroll events here and responderIgnoreScroll indicates we
|
|
// will send topTouchCancel to handle canceling touch events instead
|
|
((topLevelType === TOP_SCROLL && !nativeEvent.responderIgnoreScroll) ||
|
|
(trackedTouchCount > 0 && topLevelType === TOP_SELECTION_CHANGE) ||
|
|
isStartish(topLevelType) ||
|
|
isMoveish(topLevelType))
|
|
);
|
|
}
|
|
|
|
/**
|
|
* Returns whether or not this touch end event makes it such that there are no
|
|
* longer any touches that started inside of the current `responderInst`.
|
|
*
|
|
* @param {NativeEvent} nativeEvent Native touch end event.
|
|
* @return {boolean} Whether or not this touch end event ends the responder.
|
|
*/
|
|
function noResponderTouches(nativeEvent) {
|
|
const touches = nativeEvent.touches;
|
|
if (!touches || touches.length === 0) {
|
|
return true;
|
|
}
|
|
for (let i = 0; i < touches.length; i++) {
|
|
const activeTouch = touches[i];
|
|
const target = activeTouch.target;
|
|
if (target !== null && target !== undefined && target !== 0) {
|
|
// Is the original touch location inside of the current responder?
|
|
const targetInst = getInstanceFromNode(target);
|
|
if (isAncestor(responderInst, targetInst)) {
|
|
return false;
|
|
}
|
|
}
|
|
}
|
|
return true;
|
|
}
|
|
|
|
const ResponderEventPlugin = {
|
|
/* For unit testing only */
|
|
_getResponder: function() {
|
|
return responderInst;
|
|
},
|
|
|
|
eventTypes: eventTypes,
|
|
|
|
/**
|
|
* We must be resilient to `targetInst` being `null` on `touchMove` or
|
|
* `touchEnd`. On certain platforms, this means that a native scroll has
|
|
* assumed control and the original touch targets are destroyed.
|
|
*/
|
|
extractEvents: function(
|
|
topLevelType,
|
|
eventSystemFlags,
|
|
targetInst,
|
|
nativeEvent,
|
|
nativeEventTarget,
|
|
) {
|
|
if (isStartish(topLevelType)) {
|
|
trackedTouchCount += 1;
|
|
} else if (isEndish(topLevelType)) {
|
|
if (trackedTouchCount >= 0) {
|
|
trackedTouchCount -= 1;
|
|
} else {
|
|
console.warn(
|
|
'Ended a touch event which was not counted in `trackedTouchCount`.',
|
|
);
|
|
return null;
|
|
}
|
|
}
|
|
|
|
ResponderTouchHistoryStore.recordTouchTrack(topLevelType, nativeEvent);
|
|
|
|
let extracted = canTriggerTransfer(topLevelType, targetInst, nativeEvent)
|
|
? setResponderAndExtractTransfer(
|
|
topLevelType,
|
|
targetInst,
|
|
nativeEvent,
|
|
nativeEventTarget,
|
|
)
|
|
: null;
|
|
// Responder may or may not have transferred on a new touch start/move.
|
|
// Regardless, whoever is the responder after any potential transfer, we
|
|
// direct all touch start/move/ends to them in the form of
|
|
// `onResponderMove/Start/End`. These will be called for *every* additional
|
|
// finger that move/start/end, dispatched directly to whoever is the
|
|
// current responder at that moment, until the responder is "released".
|
|
//
|
|
// These multiple individual change touch events are are always bookended
|
|
// by `onResponderGrant`, and one of
|
|
// (`onResponderRelease/onResponderTerminate`).
|
|
const isResponderTouchStart = responderInst && isStartish(topLevelType);
|
|
const isResponderTouchMove = responderInst && isMoveish(topLevelType);
|
|
const isResponderTouchEnd = responderInst && isEndish(topLevelType);
|
|
const incrementalTouch = isResponderTouchStart
|
|
? eventTypes.responderStart
|
|
: isResponderTouchMove
|
|
? eventTypes.responderMove
|
|
: isResponderTouchEnd
|
|
? eventTypes.responderEnd
|
|
: null;
|
|
|
|
if (incrementalTouch) {
|
|
const gesture = ResponderSyntheticEvent.getPooled(
|
|
incrementalTouch,
|
|
responderInst,
|
|
nativeEvent,
|
|
nativeEventTarget,
|
|
);
|
|
gesture.touchHistory = ResponderTouchHistoryStore.touchHistory;
|
|
accumulateDirectDispatches(gesture);
|
|
extracted = accumulate(extracted, gesture);
|
|
}
|
|
|
|
const isResponderTerminate =
|
|
responderInst && topLevelType === TOP_TOUCH_CANCEL;
|
|
const isResponderRelease =
|
|
responderInst &&
|
|
!isResponderTerminate &&
|
|
isEndish(topLevelType) &&
|
|
noResponderTouches(nativeEvent);
|
|
const finalTouch = isResponderTerminate
|
|
? eventTypes.responderTerminate
|
|
: isResponderRelease
|
|
? eventTypes.responderRelease
|
|
: null;
|
|
if (finalTouch) {
|
|
const finalEvent = ResponderSyntheticEvent.getPooled(
|
|
finalTouch,
|
|
responderInst,
|
|
nativeEvent,
|
|
nativeEventTarget,
|
|
);
|
|
finalEvent.touchHistory = ResponderTouchHistoryStore.touchHistory;
|
|
accumulateDirectDispatches(finalEvent);
|
|
extracted = accumulate(extracted, finalEvent);
|
|
changeResponder(null);
|
|
}
|
|
|
|
return extracted;
|
|
},
|
|
|
|
GlobalResponderHandler: null,
|
|
|
|
injection: {
|
|
/**
|
|
* @param {{onChange: (ReactID, ReactID) => void} GlobalResponderHandler
|
|
* Object that handles any change in responder. Use this to inject
|
|
* integration with an existing touch handling system etc.
|
|
*/
|
|
injectGlobalResponderHandler(GlobalResponderHandler) {
|
|
ResponderEventPlugin.GlobalResponderHandler = GlobalResponderHandler;
|
|
},
|
|
},
|
|
};
|
|
|
|
export default ResponderEventPlugin;
|