Share via


Catalog.ListColumns Method

Definition

Overloads

ListColumns(String)

Returns a list of columns for the given table/view or temporary view. The DataFrame includes the name, description, dataType, whether it is nullable or if it is partitioned and if it is broken in buckets.

ListColumns(String, String)

Returns a list of columns for the given table/view in the specified database. The DataFrame includes the name, description, dataType, whether it is nullable or if it is partitioned and if it is broken in buckets.

ListColumns(String)

Returns a list of columns for the given table/view or temporary view. The DataFrame includes the name, description, dataType, whether it is nullable or if it is partitioned and if it is broken in buckets.

public Microsoft.Spark.Sql.DataFrame ListColumns (string tableName);
member this.ListColumns : string -> Microsoft.Spark.Sql.DataFrame
Public Function ListColumns (tableName As String) As DataFrame

Parameters

tableName
String

Is either a qualified or unqualified name that designates a table. If no database identifier is provided, it refers to a table in the current database.

Returns

DataFrame with the name, description, dataType, whether each column is nullable, if the column is partitioned and if the column is broken in buckets.

Applies to

ListColumns(String, String)

Returns a list of columns for the given table/view in the specified database. The DataFrame includes the name, description, dataType, whether it is nullable or if it is partitioned and if it is broken in buckets.

public Microsoft.Spark.Sql.DataFrame ListColumns (string dbName, string tableName);
member this.ListColumns : string * string -> Microsoft.Spark.Sql.DataFrame
Public Function ListColumns (dbName As String, tableName As String) As DataFrame

Parameters

dbName
String

Is a name that designates a database.

tableName
String

Is an unqualified name that designates a table in the specified database.

Returns

DataFrame with the name, description, dataType, whether each column is nullable, if the column is partitioned and if the column is broken in buckets.

Applies to