Functional Map API with Embedded Infinispan

What You Will Learn

How to use the Infinispan functional map API, which provides a lambda-based approach to cache operations with write-only, read-only, and read-write map views.

Prerequisites

Java 17+

Step 1: Add the Embedded Dependency

Add Infinispan core to your pom.xml:

<dependency>
   <groupId>org.infinispan</groupId>
   <artifactId>infinispan-core</artifactId>
</dependency>

Step 2: Create the Functional Map

Obtain a FunctionalMapImpl from an AdvancedCache, then create typed map views:

      cacheManager = new DefaultCacheManager();
      cacheManager.defineConfiguration("local", new ConfigurationBuilder().build());
      cache = cacheManager.<String, String>getCache("local").getAdvancedCache();
      functionalMap = FunctionalMapImpl.create(cache);
      writeOnlyMap = WriteOnlyMapImpl.create(functionalMap);
      readOnlyMap = ReadOnlyMapImpl.create(functionalMap);

Step 3: Write and Read Entries

Use write-only operations to store values and chain read-only operations with CompletableFuture:

      // Execute two parallel write-only operation to store key/value pairs
      CompletableFuture<Void> writeFuture1 = writeOnlyMap.eval("key1", "value1",
         (v, writeView) -> writeView.set(v));
      CompletableFuture<Void> writeFuture2 = writeOnlyMap.eval("key2", "value2",
         (v, writeView) -> writeView.set(v));

      // When each write-only operation completes, execute a read-only operation to retrieve the value
      CompletableFuture<String> readFuture1 =
         writeFuture1.thenCompose(r -> readOnlyMap.eval("key1", EntryView.ReadEntryView::get));
      CompletableFuture<String> readFuture2 =
         writeFuture2.thenCompose(r -> readOnlyMap.eval("key2", EntryView.ReadEntryView::get));

      // When the read-only operation completes, print it out
      System.out.printf("Created entries: %n");
      CompletableFuture<Void> end = readFuture1.thenAcceptBoth(readFuture2, (v1, v2) ->
         System.out.printf("key1 = %s%nkey2 = %s%n", v1, v2));

      // Wait for this read/write combination to finish
      end.get();

Step 4: Use Read-Write Operations

The read-write map supports atomic operations that read the previous value and write a new one, with optional metadata like lifespan:

      // Create a read-write map
      FunctionalMap.ReadWriteMap<String, String> readWriteMap = ReadWriteMapImpl.create(functionalMap);

      // Use read-write multi-key based operation to write new values
      // together with lifespan and return previous values
      Map<String, String> data = new HashMap<>();
      data.put("key1", "newValue1");
      data.put("key2", "newValue2");
      Traversable<String> previousValues = readWriteMap.evalMany(data, (v, readWriteView) -> {
         String prev = readWriteView.find().orElse(null);
         readWriteView.set(v, new MetaLifespan(Duration.ofHours(1).toMillis()));
         return prev;
      });

Step 5: Run the Tutorial

mvn package exec:java

The output shows created entries, updated entries with their new values, and the previous values before the update.

What’s Next

Try cache listeners to react to cache changes
Explore distributed streams for parallel processing
Use JCache API for a standard caching interface