szehon-ho commented on code in PR #53508:
URL: https://github.com/apache/spark/pull/53508#discussion_r2632375618
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/rules.scala:
##########
@@ -695,19 +695,36 @@ object ViewSyncSchemaToMetaStore extends (LogicalPlan =>
Unit) {
(field.dataType != planField.dataType ||
field.nullable != planField.nullable ||
(viewSchemaMode == SchemaEvolution && (
- field.getComment() != planField.getComment() ||
- field.name != planField.name)))
+ field.name != planField.name ||
+ // Only trigger redo on comment changes if preserve flag is
disabled.
+
(!session.sessionState.conf.viewSchemaEvolutionPreserveUserComments &&
+ field.getComment() != planField.getComment()))))
}
+ lazy val viewFieldsByName = viewFields.map(f => f.name -> f).toMap
+
if (redo) {
val newSchema = if (viewSchemaMode == SchemaTypeEvolution) {
val newFields = viewQuery.schema.map {
case StructField(name, dataType, nullable, _) =>
StructField(name, dataType, nullable,
- viewFields.find(_.name == name).get.metadata)
+ viewFieldsByName(name).metadata)
+ }
Review Comment:
Clarify this comment:, a nested if (as its logically under the
schemaEvolution condition). It makes more sense to me this way, but up to you.
```
val newSchema =
if (viewSchemaMode == SchemaTypeEvolution) {
val newFields = viewQuery.schema.map { planField =>
val metadata =
if (viewSchemaEvolutionPreserveUserComments) {
// Preserve old comment
viewFieldsByName.get(planField.name) match {
case Some(viewField) =>
val builder = new
MetadataBuilder().withMetadata(planField.metadata)
viewField.getComment() match {
case Some(comment) => builder.putString("comment", comment)
case None => builder.remove("comment")
}
builder.build()
case None =>
// New column, use new plan metadata
planField.metadata
}
} else {
viewFieldsByName(planField.name).metadata
}
StructField(planField.name, planField.dataType, planField.nullable,
metadata)
}
StructType(newFields)
}
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
