[SPARK-50770][SS] Removing package scope for transformWithState opera…

…tor APIs ### What changes were proposed in this pull request? Removing package scope for transformWithState operator APIs This enables us to release the transformWithState APIs from the next Spark release. While traits and methods expand visibility, the usage is blocked due to a flag. The flag will be flipped in separate PR and corresponding behavior change ticket will be filed with it. ### Why are the changes needed? To allow use of the transformWithState operator ### Does this PR introduce _any_ user-facing change? Yes ### How was this patch tested? Existing unit tests ### Was this patch authored or co-authored using generative AI tooling? No Closes #49417 from ericm-db/enable-tws. Authored-by: Eric Marnadi <[email protected]> Signed-off-by: Jungtaek Lim <[email protected]> (cherry picked from commit be77e1e) Signed-off-by: Jungtaek Lim <[email protected]>
apache · Jan 24, 2025 · f264560 · f264560
1 parent f2765f4
commit f264560
Show file tree

Hide file tree

Showing 15 changed files with 22 additions and 42 deletions.
diff --git a/sql/api/src/main/java/org/apache/spark/api/java/function/FlatMapGroupsWithStateFunction.java b/sql/api/src/main/java/org/apache/spark/api/java/function/FlatMapGroupsWithStateFunction.java
@@ -21,7 +21,6 @@
 import java.util.Iterator;
 
 import org.apache.spark.annotation.Evolving;
-import org.apache.spark.annotation.Experimental;
 import org.apache.spark.sql.streaming.GroupState;
 
 /**
@@ -32,7 +31,6 @@
  * org.apache.spark.sql.Encoder, org.apache.spark.sql.Encoder)}
  * @since 2.1.1
  */
-@Experimental
 @Evolving
 public interface FlatMapGroupsWithStateFunction<K, V, S, R> extends Serializable {
   Iterator<R> call(K key, Iterator<V> values, GroupState<S> state) throws Exception;

diff --git a/sql/api/src/main/java/org/apache/spark/api/java/function/MapGroupsWithStateFunction.java b/sql/api/src/main/java/org/apache/spark/api/java/function/MapGroupsWithStateFunction.java
@@ -21,7 +21,6 @@
 import java.util.Iterator;
 
 import org.apache.spark.annotation.Evolving;
-import org.apache.spark.annotation.Experimental;
 import org.apache.spark.sql.streaming.GroupState;
 
 /**
@@ -31,7 +30,6 @@
  * MapGroupsWithStateFunction, org.apache.spark.sql.Encoder, org.apache.spark.sql.Encoder)}
  * @since 2.1.1
  */
-@Experimental
 @Evolving
 public interface MapGroupsWithStateFunction<K, V, S, R> extends Serializable {
   R call(K key, Iterator<V> values, GroupState<S> state) throws Exception;

diff --git a/sql/api/src/main/java/org/apache/spark/sql/streaming/GroupStateTimeout.java b/sql/api/src/main/java/org/apache/spark/sql/streaming/GroupStateTimeout.java
@@ -18,7 +18,6 @@
 package org.apache.spark.sql.streaming;
 
 import org.apache.spark.annotation.Evolving;
-import org.apache.spark.annotation.Experimental;
 import org.apache.spark.sql.catalyst.plans.logical.*;
 
 /**
@@ -29,7 +28,6 @@
  *
  * @since 2.2.0
  */
-@Experimental
 @Evolving
 public class GroupStateTimeout {
   // NOTE: if you're adding new type of timeout, you should also fix the places below:

diff --git a/sql/api/src/main/java/org/apache/spark/sql/streaming/TimeMode.java b/sql/api/src/main/java/org/apache/spark/sql/streaming/TimeMode.java
@@ -18,7 +18,6 @@
 package org.apache.spark.sql.streaming;
 
 import org.apache.spark.annotation.Evolving;
-import org.apache.spark.annotation.Experimental;
 import org.apache.spark.sql.catalyst.plans.logical.EventTime$;
 import org.apache.spark.sql.catalyst.plans.logical.NoTime$;
 import org.apache.spark.sql.catalyst.plans.logical.ProcessingTime$;
@@ -27,7 +26,6 @@
  * Represents the time modes (used for specifying timers and ttl) possible for
  * the Dataset operations {@code transformWithState}.
  */
-@Experimental
 @Evolving
 public class TimeMode {
 

diff --git a/sql/api/src/main/scala/org/apache/spark/sql/streaming/ExpiredTimerInfo.scala b/sql/api/src/main/scala/org/apache/spark/sql/streaming/ExpiredTimerInfo.scala
@@ -19,14 +19,13 @@ package org.apache.spark.sql.streaming
 
 import java.io.Serializable
 
-import org.apache.spark.annotation.{Evolving, Experimental}
+import org.apache.spark.annotation.Evolving
 
 /**
  * Class used to provide access to expired timer's expiry time.
  */
-@Experimental
 @Evolving
-private[sql] trait ExpiredTimerInfo extends Serializable {
+trait ExpiredTimerInfo extends Serializable {
 
   /**
    * Get the expired timer's expiry time as milliseconds in epoch time.

diff --git a/sql/api/src/main/scala/org/apache/spark/sql/streaming/GroupState.scala b/sql/api/src/main/scala/org/apache/spark/sql/streaming/GroupState.scala
@@ -17,7 +17,7 @@
 
 package org.apache.spark.sql.streaming
 
-import org.apache.spark.annotation.{Evolving, Experimental}
+import org.apache.spark.annotation.Evolving
 import org.apache.spark.sql.catalyst.plans.logical.LogicalGroupState
 
 /**
@@ -196,7 +196,6 @@ import org.apache.spark.sql.catalyst.plans.logical.LogicalGroupState
  *   types (see `Encoder` for more details).
  * @since 2.2.0
  */
-@Experimental
 @Evolving
 trait GroupState[S] extends LogicalGroupState[S] {
 

diff --git a/sql/api/src/main/scala/org/apache/spark/sql/streaming/ListState.scala b/sql/api/src/main/scala/org/apache/spark/sql/streaming/ListState.scala
@@ -16,14 +16,13 @@
  */
 package org.apache.spark.sql.streaming
 
-import org.apache.spark.annotation.{Evolving, Experimental}
+import org.apache.spark.annotation.Evolving
 
-@Experimental
 @Evolving
 /**
  * Interface used for arbitrary stateful operations with the v2 API to capture list value state.
  */
-private[sql] trait ListState[S] extends Serializable {
+trait ListState[S] extends Serializable {
 
   /** Whether state exists or not. */
   def exists(): Boolean

diff --git a/sql/api/src/main/scala/org/apache/spark/sql/streaming/MapState.scala b/sql/api/src/main/scala/org/apache/spark/sql/streaming/MapState.scala
@@ -16,9 +16,8 @@
  */
 package org.apache.spark.sql.streaming
 
-import org.apache.spark.annotation.{Evolving, Experimental}
+import org.apache.spark.annotation.Evolving
 
-@Experimental
 @Evolving
 /**
  * Interface used for arbitrary stateful operations with the v2 API to capture map value state.

diff --git a/sql/api/src/main/scala/org/apache/spark/sql/streaming/QueryInfo.scala b/sql/api/src/main/scala/org/apache/spark/sql/streaming/QueryInfo.scala
@@ -19,15 +19,14 @@ package org.apache.spark.sql.streaming
 import java.io.Serializable
 import java.util.UUID
 
-import org.apache.spark.annotation.{Evolving, Experimental}
+import org.apache.spark.annotation.Evolving
 
 /**
  * Represents the query info provided to the stateful processor used in the arbitrary state API v2
  * to easily identify task retries on the same partition.
  */
-@Experimental
 @Evolving
-private[sql] trait QueryInfo extends Serializable {
+trait QueryInfo extends Serializable {
 
   /** Returns the streaming query id associated with stateful operator */
   def getQueryId: UUID

diff --git a/sql/api/src/main/scala/org/apache/spark/sql/streaming/StatefulProcessor.scala b/sql/api/src/main/scala/org/apache/spark/sql/streaming/StatefulProcessor.scala
@@ -19,7 +19,7 @@ package org.apache.spark.sql.streaming
 
 import java.io.Serializable
 
-import org.apache.spark.annotation.{Evolving, Experimental}
+import org.apache.spark.annotation.Evolving
 import org.apache.spark.sql.api.EncoderImplicits
 import org.apache.spark.sql.errors.ExecutionErrors
 
@@ -30,9 +30,8 @@ import org.apache.spark.sql.errors.ExecutionErrors
  * Users can also explicitly use `import implicits._` to access the EncoderImplicits and use the
  * state variable APIs relying on implicit encoders.
  */
-@Experimental
 @Evolving
-private[sql] abstract class StatefulProcessor[K, I, O] extends Serializable {
+abstract class StatefulProcessor[K, I, O] extends Serializable {
 
   // scalastyle:off
   // Disable style checker so "implicits" object can start with lowercase i
@@ -123,10 +122,8 @@ private[sql] abstract class StatefulProcessor[K, I, O] extends Serializable {
  * initial state to be initialized in the first batch. This can be used for starting a new
  * streaming query with existing state from a previous streaming query.
  */
-@Experimental
 @Evolving
-private[sql] abstract class StatefulProcessorWithInitialState[K, I, O, S]
-    extends StatefulProcessor[K, I, O] {
+abstract class StatefulProcessorWithInitialState[K, I, O, S] extends StatefulProcessor[K, I, O] {
 
   /**
    * Function that will be invoked only in the first batch for users to process initial states.

diff --git a/sql/api/src/main/scala/org/apache/spark/sql/streaming/StatefulProcessorHandle.scala b/sql/api/src/main/scala/org/apache/spark/sql/streaming/StatefulProcessorHandle.scala
@@ -18,16 +18,15 @@ package org.apache.spark.sql.streaming
 
 import java.io.Serializable
 
-import org.apache.spark.annotation.{Evolving, Experimental}
+import org.apache.spark.annotation.Evolving
 import org.apache.spark.sql.Encoder
 
 /**
  * Represents the operation handle provided to the stateful processor used in the arbitrary state
  * API v2.
  */
-@Experimental
 @Evolving
-private[sql] trait StatefulProcessorHandle extends Serializable {
+trait StatefulProcessorHandle extends Serializable {
 
   /**
    * Function to create new or return existing single value state variable of given type with ttl.

diff --git a/sql/api/src/main/scala/org/apache/spark/sql/streaming/TimerValues.scala b/sql/api/src/main/scala/org/apache/spark/sql/streaming/TimerValues.scala
@@ -19,15 +19,14 @@ package org.apache.spark.sql.streaming
 
 import java.io.Serializable
 
-import org.apache.spark.annotation.{Evolving, Experimental}
+import org.apache.spark.annotation.Evolving
 
 /**
  * Class used to provide access to timer values for processing and event time populated before
  * method invocations using the arbitrary state API v2.
  */
-@Experimental
 @Evolving
-private[sql] trait TimerValues extends Serializable {
+trait TimerValues extends Serializable {
 
   /**
    * Get the current processing time as milliseconds in epoch time.

diff --git a/sql/api/src/main/scala/org/apache/spark/sql/streaming/ValueState.scala b/sql/api/src/main/scala/org/apache/spark/sql/streaming/ValueState.scala
@@ -19,14 +19,13 @@ package org.apache.spark.sql.streaming
 
 import java.io.Serializable
 
-import org.apache.spark.annotation.{Evolving, Experimental}
+import org.apache.spark.annotation.Evolving
 
-@Experimental
 @Evolving
 /**
  * Interface used for arbitrary stateful operations with the v2 API to capture single value state.
  */
-private[sql] trait ValueState[S] extends Serializable {
+trait ValueState[S] extends Serializable {
 
   /** Whether state exists or not. */
   def exists(): Boolean

diff --git a/sql/core/src/main/scala/org/apache/spark/sql/KeyValueGroupedDataset.scala b/sql/core/src/main/scala/org/apache/spark/sql/KeyValueGroupedDataset.scala
@@ -201,7 +201,7 @@ class KeyValueGroupedDataset[K, V] private[sql](
   }
 
   /** @inheritdoc */
-  private[sql] def transformWithState[U: Encoder](
+  def transformWithState[U: Encoder](
       statefulProcessor: StatefulProcessor[K, V, U],
       timeMode: TimeMode,
       outputMode: OutputMode): Dataset[U] = {
@@ -219,7 +219,7 @@ class KeyValueGroupedDataset[K, V] private[sql](
   }
 
   /** @inheritdoc */
-  private[sql] def transformWithState[U: Encoder](
+  def transformWithState[U: Encoder](
       statefulProcessor: StatefulProcessor[K, V, U],
       eventTimeColumnName: String,
       outputMode: OutputMode): Dataset[U] = {
@@ -235,7 +235,7 @@ class KeyValueGroupedDataset[K, V] private[sql](
   }
 
   /** @inheritdoc */
-  private[sql] def transformWithState[U: Encoder, S: Encoder](
+  def transformWithState[U: Encoder, S: Encoder](
       statefulProcessor: StatefulProcessorWithInitialState[K, V, U, S],
       timeMode: TimeMode,
       outputMode: OutputMode,
@@ -257,7 +257,7 @@ class KeyValueGroupedDataset[K, V] private[sql](
   }
 
   /** @inheritdoc */
-  private[sql] def transformWithState[U: Encoder, S: Encoder](
+  def transformWithState[U: Encoder, S: Encoder](
       statefulProcessor: StatefulProcessorWithInitialState[K, V, U, S],
       eventTimeColumnName: String,
       outputMode: OutputMode,

diff --git a/sql/core/src/main/scala/org/apache/spark/sql/streaming/TestGroupState.scala b/sql/core/src/main/scala/org/apache/spark/sql/streaming/TestGroupState.scala
@@ -17,7 +17,7 @@
 
 package org.apache.spark.sql.streaming
 
-import org.apache.spark.annotation.{Evolving, Experimental}
+import org.apache.spark.annotation.Evolving
 import org.apache.spark.api.java.Optional
 import org.apache.spark.sql.execution.streaming.GroupStateImpl
 import org.apache.spark.sql.execution.streaming.GroupStateImpl._
@@ -114,7 +114,6 @@ import org.apache.spark.sql.execution.streaming.GroupStateImpl._
  *           Spark SQL types (see `Encoder` for more details).
  * @since 3.2.0
  */
-@Experimental
 @Evolving
 trait TestGroupState[S] extends GroupState[S] {
   /** Whether the state has been marked for removing */