[Misc] Refactor Module.cached extension to accept lambda instead of FunctionalInterface

This commit is contained in:
Dmitry Savvinov
2019-06-05 12:57:39 +03:00
parent b477184a3c
commit 54cdd3bfef
6 changed files with 111 additions and 87 deletions
@@ -0,0 +1,62 @@
/*
* Copyright 2010-2019 JetBrains s.r.o. and Kotlin Programming Language contributors.
* Use of this source code is governed by the Apache 2.0 license that can be found in the license/LICENSE.txt file.
*/
package org.jetbrains.kotlin.caches.project
import com.intellij.openapi.module.Module
import com.intellij.openapi.project.Project
import com.intellij.openapi.roots.ProjectRootModificationTracker
import com.intellij.openapi.util.UserDataHolder
import com.intellij.psi.util.CachedValueProvider
import com.intellij.psi.util.CachedValuesManager
fun <T> Module.cacheByClass(classForKey: Class<*>, vararg dependencies: Any, provider: () -> T): T {
return CachedValuesManager.getManager(project).cache(this, dependencies, classForKey, provider)
}
fun <T> Module.cacheByClassInvalidatingOnRootModifications(classForKey: Class<*>, provider: () -> T): T {
return cacheByClass(classForKey, ProjectRootModificationTracker.getInstance(project), provider = provider)
}
/**
* Note that it uses lambda's class for caching (essentially, anonymous class), which means that all invocations will be cached
* by the one and the same key.
* It is encouraged to use explicit class, just for the sake of readability.
*/
fun <T> Module.cacheInvalidatingOnRootModifications(provider: () -> T): T {
return cacheByClassInvalidatingOnRootModifications(provider::class.java, provider)
}
fun <T> Project.cacheByClass(classForKey: Class<*>, vararg dependencies: Any, provider: () -> T): T {
return CachedValuesManager.getManager(this).cache(this, dependencies, classForKey, provider)
}
fun <T> Project.cacheByClassInvalidatingOnRootModifications(classForKey: Class<*>, provider: () -> T): T {
return cacheByClass(classForKey, ProjectRootModificationTracker.getInstance(this), provider = provider)
}
/**
* Note that it uses lambda's class for caching (essentially, anonymous class), which means that all invocations will be cached
* by the one and the same key.
* It is encouraged to use explicit class, just for the sake of readability.
*/
fun <T> Project.cacheInvalidatingOnRootModifications(provider: () -> T): T {
return cacheByClassInvalidatingOnRootModifications(provider::class.java, provider)
}
private fun <T> CachedValuesManager.cache(
holder: UserDataHolder,
dependencies: Array<out Any>,
classForKey: Class<*>,
provider: () -> T
): T {
return getCachedValue(
holder,
getKeyForClass(classForKey),
{ CachedValueProvider.Result.create(provider(), dependencies) },
false
)
}