Add comments to async query handlers

spark/src/main/java/org/opensearch/sql/spark/dispatcher/BatchQueryHandler.java

@@ -28,6 +28,10 @@
 import org.opensearch.sql.spark.leasemanager.model.LeaseRequest;
 import org.opensearch.sql.spark.response.JobExecutionResponseReader;
 
+/**
+ * The handler for batch query. With batch query, queries are executed as single batch. The queries
+ * are sent along with job execution request ({@link StartJobRequest}) to spark.
+ */
 @RequiredArgsConstructor
 public class BatchQueryHandler extends AsyncQueryHandler {
   private final EMRServerlessClient emrServerlessClient;

spark/src/main/java/org/opensearch/sql/spark/dispatcher/IndexDMLHandler.java

@@ -31,7 +31,12 @@
 import org.opensearch.sql.spark.flint.operation.FlintIndexOpFactory;
 import org.opensearch.sql.spark.response.JobExecutionResponseReader;
 
-/** Handle Index DML query. includes * DROP * ALT? */
+/**
+ * The handler for Index DML (Data Manipulation Language) query. Handles DROP/ALTER/VACUUM operation
+ * for flint indices. It will stop streaming query job as needed (e.g. when the flint index is
+ * automatically updated by a streaming query, the streaming query is stopped when the index is
+ * dropped)
+ */
 @RequiredArgsConstructor
 public class IndexDMLHandler extends AsyncQueryHandler {
   private static final Logger LOG = LogManager.getLogger();

spark/src/main/java/org/opensearch/sql/spark/dispatcher/InteractiveQueryHandler.java

@@ -35,6 +35,12 @@
 import org.opensearch.sql.spark.leasemanager.model.LeaseRequest;
 import org.opensearch.sql.spark.response.JobExecutionResponseReader;
 
+/**
+ * The handler for interactive query. With interactive query, a session will be first established
+ * and then the session will be reused for the following queries(statements). Session is an
+ * abstraction of spark job, and once the job is started, the job will continuously poll the
+ * statements and execute query specified in it.
+ */
 @RequiredArgsConstructor
 public class InteractiveQueryHandler extends AsyncQueryHandler {
   private final SessionManager sessionManager;

spark/src/main/java/org/opensearch/sql/spark/dispatcher/RefreshQueryHandler.java

@@ -20,7 +20,10 @@
 import org.opensearch.sql.spark.leasemanager.LeaseManager;
 import org.opensearch.sql.spark.response.JobExecutionResponseReader;
 
-/** Handle Refresh Query. */
+/**
+ * The handler for refresh query. Refresh query is one time query request to refresh(update) flint
+ * index, and new job is submitted to Spark.
+ */
 public class RefreshQueryHandler extends BatchQueryHandler {
 
   private final FlintIndexMetadataService flintIndexMetadataService;

spark/src/main/java/org/opensearch/sql/spark/dispatcher/StreamingQueryHandler.java

@@ -26,7 +26,10 @@
 import org.opensearch.sql.spark.leasemanager.model.LeaseRequest;
 import org.opensearch.sql.spark.response.JobExecutionResponseReader;
 
-/** Handle Streaming Query. */
+/**
+ * The handler for streaming query. Streaming query is a job to continuously update flint index.
+ * Once started, the job can be stopped by IndexDML query.
+ */
 public class StreamingQueryHandler extends BatchQueryHandler {
   private final EMRServerlessClient emrServerlessClient;
 

spark/src/main/java/org/opensearch/sql/spark/flint/IndexDMLResultStorageService.java

@@ -7,6 +7,9 @@
 
 import org.opensearch.sql.spark.dispatcher.model.IndexDMLResult;
 
+/**
+ * Abstraction over the IndexDMLResult storage. It stores the result of IndexDML query execution.
+ */
 public interface IndexDMLResultStorageService {
   IndexDMLResult createIndexDMLResult(IndexDMLResult result, String datasourceName);
 }