This is an automated email from the ASF dual-hosted git repository.
schang pushed a commit to branch branch-0.x
in repository https://gitbox.apache.org/repos/asf/hudi.git
The following commit(s) were added to refs/heads/branch-0.x by this push:
new 7f9b469b23cb fix: Databricks Spark 3.4 Runtime compatibility for
reading Hudi tables (#18291)
7f9b469b23cb is described below
commit 7f9b469b23cb604531a5a04ac4225478aced6773
Author: Y Ethan Guo <[email protected]>
AuthorDate: Sat Mar 7 21:55:21 2026 -0800
fix: Databricks Spark 3.4 Runtime compatibility for reading Hudi tables
(#18291)
---
.../org/apache/hudi/DatabricksRuntimeHelper.scala | 75 ++++++++++++++++++++++
.../HoodieSpark34PartitionedFileUtils.scala | 3 +-
2 files changed, 77 insertions(+), 1 deletion(-)
diff --git
a/hudi-spark-datasource/hudi-spark-common/src/main/scala/org/apache/hudi/DatabricksRuntimeHelper.scala
b/hudi-spark-datasource/hudi-spark-common/src/main/scala/org/apache/hudi/DatabricksRuntimeHelper.scala
new file mode 100644
index 000000000000..2ddd07ceb0bf
--- /dev/null
+++
b/hudi-spark-datasource/hudi-spark-common/src/main/scala/org/apache/hudi/DatabricksRuntimeHelper.scala
@@ -0,0 +1,75 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing,
+ * software distributed under the License is distributed on an
+ * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+ * KIND, either express or implied. See the License for the
+ * specific language governing permissions and limitations
+ * under the License.
+ */
+
+package org.apache.hudi
+
+import org.apache.hadoop.fs.FileStatus
+import org.apache.spark.sql.catalyst.InternalRow
+import org.apache.spark.sql.execution.datasources.PartitionDirectory
+
+import scala.util.Try
+
+/**
+ * Compatibility helper for Databricks Spark runtime.
+ */
+object DatabricksRuntimeHelper {
+ private lazy val fileStatusWithMetadataClass: Option[Class[_]] =
+
Try(Class.forName("org.apache.spark.sql.execution.datasources.FileStatusWithMetadata")).toOption
+ private lazy val fileStatusWithMetadataModuleClass =
+
Try(Class.forName("org.apache.spark.sql.execution.datasources.FileStatusWithMetadata$")).toOption
+ private lazy val fileStatusWithMetadataModule =
fileStatusWithMetadataModuleClass.map { cls =>
+ cls.getField("MODULE$").get(null) // the singleton companion object
instance
+ }
+ private lazy val fileStatusWithMetadataInstantiationMethod =
fileStatusWithMetadataClass.flatMap { cls =>
+ Try(cls.getMethod("apply", classOf[FileStatus])).toOption
+ }
+
+ private lazy val partitionDirectoryConstructor:
Option[java.lang.reflect.Constructor[_]] =
+ fileStatusWithMetadataClass.flatMap { _ =>
+ Try(classOf[PartitionDirectory].getConstructor(classOf[InternalRow],
classOf[Seq[_]])).toOption
+ }
+
+ /**
+ * Creates a PartitionDirectory, using reflection on Databricks runtime where
+ * PartitionDirectory expects (InternalRow, Seq[FileStatusWithMetadata])
instead of
+ * (InternalRow, Seq[FileStatus]).
+ *
+ * Databricks Spark 3.4 runtime backports FileStatusWithMetadata from Spark
3.5.
+ * On Databricks Spark 3.4, PartitionDirectory expects FileStatusWithMetadata
+ * instead of plain FileStatus.
+ *
+ * FileStatusWithMetadata does NOT extend FileStatus (it wraps it via
composition),
+ * so individual element casts are impossible. We use reflection to
construct it.
+ *
+ * On Databricks, constructs FileStatusWithMetadata.apply(fileStatus) for
each file
+ * status, then reflectively constructs PartitionDirectory.
+ * On standard Spark, falls back to normal PartitionDirectory construction.
+ */
+ def newPartitionDirectory(internalRow: InternalRow,
+ statuses: Seq[FileStatus],
+ fallback: (InternalRow, Seq[FileStatus]) =>
PartitionDirectory): PartitionDirectory = {
+ (fileStatusWithMetadataInstantiationMethod, fileStatusWithMetadataModule,
partitionDirectoryConstructor) match {
+ case (Some(method), Some(module), Some(ctor)) =>
+ val wrappedStatuses = statuses.map(e => method.invoke(module, e))
+ ctor.newInstance(internalRow,
wrappedStatuses).asInstanceOf[PartitionDirectory]
+ case _ =>
+ fallback(internalRow, statuses)
+ }
+ }
+}
diff --git
a/hudi-spark-datasource/hudi-spark3.4.x/src/main/scala/org/apache/spark/sql/execution/datasources/HoodieSpark34PartitionedFileUtils.scala
b/hudi-spark-datasource/hudi-spark3.4.x/src/main/scala/org/apache/spark/sql/execution/datasources/HoodieSpark34PartitionedFileUtils.scala
index c51e13763c76..fdc5acbc5fe7 100644
---
a/hudi-spark-datasource/hudi-spark3.4.x/src/main/scala/org/apache/spark/sql/execution/datasources/HoodieSpark34PartitionedFileUtils.scala
+++
b/hudi-spark-datasource/hudi-spark3.4.x/src/main/scala/org/apache/spark/sql/execution/datasources/HoodieSpark34PartitionedFileUtils.scala
@@ -19,6 +19,7 @@
package org.apache.spark.sql.execution.datasources
+import org.apache.hudi.DatabricksRuntimeHelper
import org.apache.hudi.storage.StoragePath
import org.apache.hadoop.fs.FileStatus
@@ -49,6 +50,6 @@ object HoodieSpark34PartitionedFileUtils extends
HoodieSparkPartitionedFileUtils
}
override def newPartitionDirectory(internalRow: InternalRow, statuses:
Seq[FileStatus]): PartitionDirectory = {
- PartitionDirectory(internalRow, statuses)
+ DatabricksRuntimeHelper.newPartitionDirectory(internalRow, statuses, (row,
files) => PartitionDirectory(row, files))
}
}
