Optimizes data-fetching by batching and eliminating duplicate calls, reducing database/API load. Utilizes coroutines for asynchronous data loading, ensuring efficient and streamlined data retrieval.
A pure, multiplatform (js and jvm) Kotlin implementation of DataLoader. This library is using Kotlin Coroutines.
A DataLoader is a mechanism to optimize data-fetching. Instead of issuing multiple calls one after the other, the DataLoader will collect all calls and execute them together as one. This alone has the huge benefit of fewer calls to your db/api, therefore reducing load. The second optimization is to eliminate duplicate calls.
Assumption: You want to fetch a list of Projects with their members. The relation between Project and Person is m:n, therefore a Person can work in multiple projects.
The data:
In a naive approach you will need 6 requests to get all your data:
The same scenario can be optimized to 2 requests with a DataLoader:
How does it work?
If you want to load the Person with id 1, you simply call dataLoader.loadAsync(1).
You won't get the result immediately, but you'll get a Deferred, which is an equivalent of a Future or Promise in other languages.
In the example above the calls will be:
dataLoader.loadAsync(1)dataLoader.loadAsync(2)dataLoader.loadAsync(2)dataLoader.loadAsync(1)dataLoader.loadAsync(3)By now no call to actually fetch the data has been made.
The actual fetch happens when you call dataLoader.dispatch().
In this case the ids 1,2 and 3 will be fetched in one call.
Since calling dataLoader.loadAsync() with the same id will result in the same Deffered being returned, no id is fetched more than once.
Currently, only jvm and js is supported.
repositories {
mavenCentral()
}
kotlin {
sourceSets {
val commonMain by getting {
dependencies {
implementation("de.nidomiro:KDataLoader:0.5.1")
}
}
}
}
For JVM-only or js only use the following instead of de.nidomiro:KDataLoader:
de.nidomiro:KDataLoader-jvmde.nidomiro:KDataLoader-js<dependency>
<groupId>de.nidomiro</groupId>
<artifactId>KDataLoader-jvm</artifactId>
<version>0.5.1</version>
</dependency>
Since this library uses Kotlin Coroutines every example assumes you are in a coroutine.
val batchLoader: BatchLoader<, > =
{ keys -> keys.map { ExecutionResult.Success(it) } }
dataLoader = dataLoader(batchLoader) {
configure {
cache = DefaultCacheImpl()
cacheEnabled =
cacheExceptions =
batchLoadEnabled =
batchSize = .MAX_VALUE
}
prime( to )
}
value1 = dataLoader.loadAsync()
value2 = dataLoader.loadAsync()
dataLoader.dispatch()
println()
println()
This Library uses Semantic Versioning.
This Library is Licensed under the MIT License.
{
"projects":[
{
"name":"Project A",
"members":[
{
"id":1,
"name":"John"
},
{
"id":2,
"name":"Anne"
}
]
},
{
"name":"Project B",
"members":[
{
"id":2,
"name":"Anne"
}
]
},
{
"name":"Project C",
"members":[
{
"id":1,
"name":"John"
},
{
"id":3,
"name":"Peter"
}
]
}
]
}
Surfaced from shared tags and platforms — no rankings paid for.