[GH-2609] Fix the thread leak issue in VertexAiTextEmbeddingModel

The PredictionServiceClient was not being closed. Connections are kept open preventing resources from being disposed properly.

Signed-off-by: Rodrigo Malara <[email protected]>
Signed-off-by: rmalara <[email protected]>

Author: rmalara

models/spring-ai-vertex-ai-embedding/src/main/java/org/springframework/ai/vertexai/embedding/text/VertexAiTextEmbeddingModel.java

@@ -128,7 +128,7 @@ public EmbeddingResponse call(EmbeddingRequest request) {
 			.observation(this.observationConvention, DEFAULT_OBSERVATION_CONVENTION, () -> observationContext,
 					this.observationRegistry)
 			.observe(() -> {
-				try(PredictionServiceClient client = createPredictionServiceClient()) {
+				try (PredictionServiceClient client = createPredictionServiceClient()) {
 
 					EndpointName endpointName = this.connectionDetails.getEndpointName(finalOptions.getModel());