InternalMutatorMutex.kt
/*
* Copyright 2020 The Android Open Source Project
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package androidx.compose.material
import androidx.compose.foundation.MutatePriority
import androidx.compose.runtime.Stable
import kotlinx.coroutines.CancellationException
import kotlinx.coroutines.Job
import kotlinx.coroutines.coroutineScope
import kotlinx.coroutines.sync.Mutex
import kotlinx.coroutines.sync.withLock
/*** This is an internal copy of androidx.compose.foundation.MutatorMutex with an additional
* tryMutate method. Do not modify, except for tryMutate. ***/
expect class InternalAtomicReference<V>(value: V) {
fun get(): V
fun set(value: V)
fun getAndSet(value: V): V
fun compareAndSet(expect: V, newValue: V): Boolean
}
/**
* Mutual exclusion for UI state mutation over time.
*
* [mutate] permits interruptible state mutation over time using a standard [MutatePriority].
* A [InternalMutatorMutex] enforces that only a single writer can be active at a time for a particular
* state resource. Instead of queueing callers that would acquire the lock like a traditional
* [Mutex], new attempts to [mutate] the guarded state will either cancel the current mutator or
* if the current mutator has a higher priority, the new caller will throw [CancellationException].
*
* [InternalMutatorMutex] should be used for implementing hoisted state objects that many mutators may
* want to manipulate over time such that those mutators can coordinate with one another. The
* [InternalMutatorMutex] instance should be hidden as an implementation detail. For example:
*
*/
@Stable
internal class InternalMutatorMutex {
private class Mutator(val priority: MutatePriority, val job: Job) {
fun canInterrupt(other: Mutator) = priority >= other.priority
fun cancel() = job.cancel()
}
private val currentMutator = InternalAtomicReference<Mutator?>(null)
private val mutex = Mutex()
private fun tryMutateOrCancel(mutator: Mutator) {
while (true) {
val oldMutator = currentMutator.get()
if (oldMutator == null || mutator.canInterrupt(oldMutator)) {
if (currentMutator.compareAndSet(oldMutator, mutator)) {
oldMutator?.cancel()
break
}
} else throw CancellationException("Current mutation had a higher priority")
}
}
/**
* Enforce that only a single caller may be active at a time.
*
* If [mutate] is called while another call to [mutate] or [mutateWith] is in progress, their
* [priority] values are compared. If the new caller has a [priority] equal to or higher than
* the call in progress, the call in progress will be cancelled, throwing
* [CancellationException] and the new caller's [block] will be invoked. If the call in
* progress had a higher [priority] than the new caller, the new caller will throw
* [CancellationException] without invoking [block].
*
* @param priority the priority of this mutation; [MutatePriority.Default] by default.
* Higher priority mutations will interrupt lower priority mutations.
* @param block mutation code to run mutually exclusive with any other call to [mutate],
* [mutateWith] or [tryMutate].
*/
suspend fun <R> mutate(
priority: MutatePriority = MutatePriority.Default,
block: suspend () -> R
) = coroutineScope {
val mutator = Mutator(priority, coroutineContext[Job]!!)
tryMutateOrCancel(mutator)
mutex.withLock {
try {
block()
} finally {
currentMutator.compareAndSet(mutator, null)
}
}
}
/**
* Enforce that only a single caller may be active at a time.
*
* If [mutateWith] is called while another call to [mutate] or [mutateWith] is in progress,
* their [priority] values are compared. If the new caller has a [priority] equal to or
* higher than the call in progress, the call in progress will be cancelled, throwing
* [CancellationException] and the new caller's [block] will be invoked. If the call in
* progress had a higher [priority] than the new caller, the new caller will throw
* [CancellationException] without invoking [block].
*
* This variant of [mutate] calls its [block] with a [receiver], removing the need to create
* an additional capturing lambda to invoke it with a receiver object. This can be used to
* expose a mutable scope to the provided [block] while leaving the rest of the state object
* read-only. For example:
*
* @param receiver the receiver `this` that [block] will be called with
* @param priority the priority of this mutation; [MutatePriority.Default] by default.
* Higher priority mutations will interrupt lower priority mutations.
* @param block mutation code to run mutually exclusive with any other call to [mutate],
* [mutateWith] or [tryMutate].
*/
suspend fun <T, R> mutateWith(
receiver: T,
priority: MutatePriority = MutatePriority.Default,
block: suspend T.() -> R
) = coroutineScope {
val mutator = Mutator(priority, coroutineContext[Job]!!)
tryMutateOrCancel(mutator)
mutex.withLock {
try {
receiver.block()
} finally {
currentMutator.compareAndSet(mutator, null)
}
}
}
/**
* Attempt to mutate synchronously if there is no other active caller.
* If there is no other active caller, the [block] will be executed in a lock. If there is
* another active caller, this method will return false, indicating that the active caller
* needs to be cancelled through a [mutate] or [mutateWith] call with an equal or higher
* mutation priority.
*
* Calls to [mutate] and [mutateWith] will suspend until execution of the [block] has finished.
*
* @param block mutation code to run mutually exclusive with any other call to [mutate],
* [mutateWith] or [tryMutate].
* @return true if the [block] was executed, false if there was another active caller and the
* [block] was not executed.
*/
fun tryMutate(block: () -> Unit): Boolean {
val didLock = mutex.tryLock()
if (didLock) {
try {
block()
} finally {
mutex.unlock()
}
}
return didLock
}
}