// non-temporary global or local view
// database defined
// externalCatalog.getTable returns non-VIEW table
// Make the example reproducible
val tableName = "t1"
spark.sharedState.externalCatalog.dropTable(
db = "default",
table = tableName,
ignoreIfNotExists = true,
purge = true)
spark.range(10).write.saveAsTable(tableName)
scala> :type spark.sessionState.catalog
org.apache.spark.sql.catalyst.catalog.SessionCatalog
import org.apache.spark.sql.catalyst.TableIdentifier
val plan = spark.sessionState.catalog.lookupRelation(TableIdentifier(tableName))
scala> println(plan.numberedTreeString)
00 'SubqueryAlias t1
01 +- 'UnresolvedCatalogRelation `default`.`t1`, org.apache.hadoop.hive.ql.io.parquet.serde.ParquetHiveSerDeUnresolvedCatalogRelation Leaf Logical Operator — Placeholder of Catalog Tables
UnresolvedCatalogRelation is a leaf logical operator that acts as a placeholder in a logical query plan for FindDataSourceTable logical evaluation rule to resolve it to a concrete relation logical operator (i.e. LogicalRelations for data source tables or HiveTableRelations for hive tables).
UnresolvedCatalogRelation is created when SessionCatalog is requested to find a relation (for DescribeTableCommand logical command or ResolveRelations logical evaluation rule).
When created, UnresolvedCatalogRelation asserts that the database is specified.
UnresolvedCatalogRelation can never be resolved and is converted to a LogicalRelation for a data source table or a HiveTableRelation for a hive table at analysis phase.
UnresolvedCatalogRelation uses an empty output schema.
UnresolvedCatalogRelation takes a single CatalogTable when created.