[GitHub] [spark] dtenedor commented on a diff in pull request #37423: [SPARK-39985][SQL] Enable implicit DEFAULT column values in inserts from DataFrames

[GitBox <gi...@apache.org>]

dtenedor commented on code in PR #37423:
URL: https://github.com/apache/spark/pull/37423#discussion_r943819176


##########
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/ResolveDefaultColumns.scala:
##########
@@ -262,16 +264,30 @@ case class ResolveDefaultColumns(catalog: SessionCatalog) extends Rule[LogicalPl
    * Updates an inline table to generate missing default column values.
    */
   private def addMissingDefaultValuesForInsertFromInlineTable(
-      table: UnresolvedInlineTable,
-      insertTableSchemaWithoutPartitionColumns: StructType): UnresolvedInlineTable = {
-    val numQueryOutputs: Int = table.rows(0).size
+      node: LogicalPlan,
+      insertTableSchemaWithoutPartitionColumns: StructType): LogicalPlan = {
+    val numQueryOutputs: Int = node match {
+      case table: UnresolvedInlineTable => table.rows(0).size
+      case local: LocalRelation => local.data(0).numFields
+    }
     val schema = insertTableSchemaWithoutPartitionColumns
     val newDefaultExpressions: Seq[Expression] =
       getDefaultExpressionsForInsert(numQueryOutputs, schema)
     val newNames: Seq[String] = schema.fields.drop(numQueryOutputs).map { _.name }
-    table.copy(
-      names = table.names ++ newNames,
-      rows = table.rows.map { row => row ++ newDefaultExpressions })
+    node match {
+      case table: UnresolvedInlineTable =>
+        table.copy(
+          names = table.names ++ newNames,
+          rows = table.rows.map { row => row ++ newDefaultExpressions })
+      case local: LocalRelation if newDefaultExpressions.nonEmpty =>
+        val colTypes = StructType(local.output.map(col => StructField(col.name, col.dataType)))
+        UnresolvedInlineTable(

Review Comment:
   I tried this a few different ways, since the `ResolveInlineTables` rule can convert the `UnresolvedInlineTable` to a `LocalRelation`. But I found this did not work well because this method adds unresolved `DEFAULT` columns and delegates replacing them to `addMissingDefaultValuesForInsertFromProject` later. I added a comment to help explain this.



-- 
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: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org