Skip to content

[SPARK-53482][SQL] MERGE INTO support for when source has less nested field than target #52347

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 2 commits into
base: master
Choose a base branch
from
Open

[SPARK-53482][SQL] MERGE INTO support for when source has less nested field than target #52347

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
f8d4f82
[SPARK-53482][SQL] MERGE INTO support for when source has less nested…
szehon-ho Sep 14, 2025
24136bf
Add test
szehon-ho Sep 15, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 3 additions & 2 deletions sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/AssignmentUtils.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -121,7 +121,7 @@ object AssignmentUtils extends SQLConfHelper with CastSupport {
val actualAttr = restoreActualType(attr)
val value = matchingAssignments.head.value
TableOutputResolver.resolveUpdate(
"", value, actualAttr, conf, err => errors += err, colPath)
"", value, actualAttr, conf, err => errors += err, colPath, true)
}
Assignment(attr, resolvedValue)
}
Expand Down Expand Up @@ -165,7 +165,8 @@ object AssignmentUtils extends SQLConfHelper with CastSupport {
TableOutputResolver.checkNullability(colExpr, col, conf, colPath)
} else if (exactAssignments.nonEmpty) {
val value = exactAssignments.head.value
TableOutputResolver.resolveUpdate("", value, col, conf, addError, colPath)
TableOutputResolver.resolveUpdate("", value, col, conf, addError, colPath,
fillDefaultValue = true)
} else {
applyFieldAssignments(col, colExpr, fieldAssignments, addError, colPath)
}
Expand Down
36 changes: 22 additions & 14 deletions sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/TableOutputResolver.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -125,7 +125,8 @@ object TableOutputResolver extends SQLConfHelper with Logging {
col: Attribute,
conf: SQLConf,
addError: String => Unit,
colPath: Seq[String]): Expression = {
colPath: Seq[String],
fillDefaultValue: Boolean): Expression = {

(value.dataType, col.dataType) match {
// no need to reorder inner fields or cast if types are already compatible
Expand All @@ -140,17 +141,17 @@ object TableOutputResolver extends SQLConfHelper with Logging {
case (valueType: StructType, colType: StructType) =>
val resolvedValue = resolveStructType(
tableName, value, valueType, col, colType,
byName = true, conf, addError, colPath)
byName = true, conf, addError, colPath, fillDefaultValue)
resolvedValue.getOrElse(value)
case (valueType: ArrayType, colType: ArrayType) =>
val resolvedValue = resolveArrayType(
tableName, value, valueType, col, colType,
byName = true, conf, addError, colPath)
byName = true, conf, addError, colPath, fillDefaultValue)
resolvedValue.getOrElse(value)
case (valueType: MapType, colType: MapType) =>
val resolvedValue = resolveMapType(
tableName, value, valueType, col, colType,
byName = true, conf, addError, colPath)
byName = true, conf, addError, colPath, fillDefaultValue)
resolvedValue.getOrElse(value)
case _ =>
checkUpdate(tableName, value, col, conf, addError, colPath)
Expand Down Expand Up @@ -301,15 +302,15 @@ object TableOutputResolver extends SQLConfHelper with Logging {
case (matchedType: StructType, expectedType: StructType) =>
resolveStructType(
tableName, matchedCol, matchedType, actualExpectedCol, expectedType,
byName = true, conf, addError, newColPath)
byName = true, conf, addError, newColPath, fillDefaultValue)
case (matchedType: ArrayType, expectedType: ArrayType) =>
resolveArrayType(
tableName, matchedCol, matchedType, actualExpectedCol, expectedType,
byName = true, conf, addError, newColPath)
byName = true, conf, addError, newColPath, fillDefaultValue)
case (matchedType: MapType, expectedType: MapType) =>
resolveMapType(
tableName, matchedCol, matchedType, actualExpectedCol, expectedType,
byName = true, conf, addError, newColPath)
byName = true, conf, addError, newColPath, fillDefaultValue)
case _ =>
checkField(
tableName, actualExpectedCol, matchedCol, byName = true, conf, addError, newColPath)
Expand Down Expand Up @@ -421,13 +422,15 @@ object TableOutputResolver extends SQLConfHelper with Logging {
byName: Boolean,
conf: SQLConf,
addError: String => Unit,
colPath: Seq[String]): Option[NamedExpression] = {
colPath: Seq[String],
fillDefaultValue: Boolean = false): Option[NamedExpression] = {
val nullCheckedInput = checkNullability(input, expected, conf, colPath)
val fields = inputType.zipWithIndex.map { case (f, i) =>
Alias(GetStructField(nullCheckedInput, i, Some(f.name)), f.name)()
}
val resolved = if (byName) {
reorderColumnsByName(tableName, fields, toAttributes(expectedType), conf, addError, colPath)
reorderColumnsByName(tableName, fields, toAttributes(expectedType), conf, addError, colPath,
fillDefaultValue)
} else {
resolveColumnsByPosition(
tableName, fields, toAttributes(expectedType), conf, addError, colPath)
Expand All @@ -454,13 +457,15 @@ object TableOutputResolver extends SQLConfHelper with Logging {
byName: Boolean,
conf: SQLConf,
addError: String => Unit,
colPath: Seq[String]): Option[NamedExpression] = {
colPath: Seq[String],
fillDefaultValue: Boolean = false): Option[NamedExpression] = {
val nullCheckedInput = checkNullability(input, expected, conf, colPath)
val param = NamedLambdaVariable("element", inputType.elementType, inputType.containsNull)
val fakeAttr =
AttributeReference("element", expectedType.elementType, expectedType.containsNull)()
val res = if (byName) {
reorderColumnsByName(tableName, Seq(param), Seq(fakeAttr), conf, addError, colPath)
reorderColumnsByName(tableName, Seq(param), Seq(fakeAttr), conf, addError, colPath,
fillDefaultValue)
} else {
resolveColumnsByPosition(tableName, Seq(param), Seq(fakeAttr), conf, addError, colPath)
}
Expand Down Expand Up @@ -488,13 +493,15 @@ object TableOutputResolver extends SQLConfHelper with Logging {
byName: Boolean,
conf: SQLConf,
addError: String => Unit,
colPath: Seq[String]): Option[NamedExpression] = {
colPath: Seq[String],
fillDefaultValue: Boolean = false): Option[NamedExpression] = {
val nullCheckedInput = checkNullability(input, expected, conf, colPath)

val keyParam = NamedLambdaVariable("key", inputType.keyType, nullable = false)
val fakeKeyAttr = AttributeReference("key", expectedType.keyType, nullable = false)()
val resKey = if (byName) {
reorderColumnsByName(tableName, Seq(keyParam), Seq(fakeKeyAttr), conf, addError, colPath)
reorderColumnsByName(tableName, Seq(keyParam), Seq(fakeKeyAttr), conf, addError, colPath,
fillDefaultValue)
} else {
resolveColumnsByPosition(tableName, Seq(keyParam), Seq(fakeKeyAttr), conf, addError, colPath)
}
Expand All @@ -504,7 +511,8 @@ object TableOutputResolver extends SQLConfHelper with Logging {
val fakeValueAttr =
AttributeReference("value", expectedType.valueType, expectedType.valueContainsNull)()
val resValue = if (byName) {
reorderColumnsByName(tableName, Seq(valueParam), Seq(fakeValueAttr), conf, addError, colPath)
reorderColumnsByName(tableName, Seq(valueParam), Seq(fakeValueAttr), conf, addError, colPath,
fillDefaultValue)
} else {
resolveColumnsByPosition(
tableName, Seq(valueParam), Seq(fakeValueAttr), conf, addError, colPath)
Expand Down
4 changes: 2 additions & 2 deletions sql/catalyst/src/test/scala/org/apache/spark/sql/connector/catalog/InMemoryBaseTable.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -892,8 +892,8 @@ private class BufferedRowsReader(
if (arrayData == null) {
null
} else {
extractArrayValue(arrayData, elementType,
readSchema.fields(readIndex).dataType)
val writeType = writeSchema.fields(writeIndex).dataType.asInstanceOf[ArrayType]
Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

fix a bug here in InMemoryDataSource schema evolution for struct within arrays

extractArrayValue(arrayData, elementType, writeType.elementType)
}

case dt =>
Expand Down
Loading