Removed downcasting and using moveToFirstToken

Author: leo-hoet

x-pack/plugin/inference/src/main/java/org/elasticsearch/xpack/inference/services/googlevertexai/GoogleVertexAiService.java

@@ -222,11 +222,11 @@ protected void doInfer(
             return;
         }
 
-        var completionModel = (GoogleVertexAiCompletionModel) model;
+        GoogleVertexAiModel googleVertexAiModel = (GoogleVertexAiModel) model;
 
         var actionCreator = new GoogleVertexAiActionCreator(getSender(), getServiceComponents());
 
-        var action = completionModel.accept(actionCreator, taskSettings);
+        var action = googleVertexAiModel.accept(actionCreator, taskSettings);
         action.execute(inputs, timeout, listener);
     }
 

x-pack/plugin/inference/src/main/java/org/elasticsearch/xpack/inference/services/googlevertexai/response/GoogleVertexAiCompletionResponseEntity.java

@@ -21,6 +21,8 @@
 import java.io.IOException;
 import java.nio.charset.StandardCharsets;
 
+import static org.elasticsearch.xpack.inference.external.response.XContentUtils.moveToFirstToken;
+
 public class GoogleVertexAiCompletionResponseEntity {
     /**
      * Parses the response from Google Vertex AI's generateContent endpoint
@@ -91,7 +93,7 @@ public static InferenceServiceResults fromResponse(Request request, HttpResult r
             XContentParser parser = XContentFactory.xContent(XContentType.JSON)
                 .createParser(XContentParserConfiguration.EMPTY, responseJson)
         ) {
-            parser.nextToken();
+            moveToFirstToken(parser);
             chunk = GoogleVertexAiUnifiedStreamingProcessor.GoogleVertexAiChatCompletionChunkParser.parse(parser);
         }
         var results = chunk.choices().stream().map(choice -> choice.delta().content()).map(ChatCompletionResults.Result::new).toList();