Course contentsShow
← Java
Lesson 2806 of 2,87051. CompletableFutureFree lesson

thenAccept: Consuming Results Without Returning a Value

Learn how thenAccept processes a CompletableFuture's result for side effects without producing a new value.

thenAccept: Consuming Results Without Returning a Value

What you'll learn: How to use thenAccept to process a CompletableFuture's result for side effects without producing a new value.

The Core Idea

While thenApply transforms a result into a new value (think mapping), thenAccept is designed for terminal operations where you want to do something with the result but don't need to return anything. It's like the difference between a function that returns a transformed value versus a void method that performs an action.

Think of it like a package delivery: thenApply is like receiving a package and repackaging it into something new. thenAccept is like receiving the package and actually using what's inside—no repackaging, just consumption.

How thenAccept Works

The thenAccept method takes a Consumer<T> (a function that accepts a value but returns nothing) and returns a CompletableFuture<Void>. This signals that the pipeline continues, but no meaningful value flows through anymore—only completion status.

CompletableFuture.supplyAsync(() -> "Hello, World!")
    .thenApply(String::toUpperCase)
    .thenAccept(result -> {
        // Side effect: print to console
        System.out.println("Result: " + result);
        // No return value needed
    });

When to Use thenAccept

Use thenAccept when you've reached the end of your transformation pipeline and need to:

  • Log results
  • Update UI components
  • Write to databases
  • Send notifications
  • Any other side-effect operation

Since it returns CompletableFuture<Void>, you can still chain further operations (like exception handling) or block on completion with join().

Key Takeaway: thenAccept is your terminal consumer for CompletableFuture pipelines—use it when you need side effects without producing new values, completing the async computation chain with an action rather than a transformation.