spark技术分享ScalarSubquery (SubqueryExpression) Expression
ScalarSubquery is a SubqueryExpression that returns a single row and a single column only.
ScalarSubquery represents a structured query that can be used as a “column”.
|
Important
|
Spark SQL uses the name of ScalarSubquery twice to represent a SubqueryExpression (this page) and an ExecSubqueryExpression. You’ve been warned.
|
ScalarSubquery is created exclusively when AstBuilder is requested to parse a subquery expression.
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 |
// FIXME DEMO // Borrowed from ExpressionParserSuite.scala // ScalarSubquery(table("tbl").select('max.function('val))) > 'current) val sql = "(select max(val) from tbl) > current" // 'a === ScalarSubquery(table("s").select('b)) val sql = "a = (select b from s)" // Borrowed from PlanParserSuite.scala // table("t").select(ScalarSubquery(table("s").select('max.function('b))).as("ss")) val sql = "select (select max(b) from s) ss from t" // table("t").where('a === ScalarSubquery(table("s").select('b))).select(star()) val sql = "select * from t where a = (select b from s)" // table("t").groupBy('g)('g).where('a > ScalarSubquery(table("s").select('b))) val sql = "select g from t group by g having a > (select b from s)" |
Creating ScalarSubquery Instance
ScalarSubquery takes the following when created:
-
Subquery logical plan
-
Child expressions (default: no children)
-
Expression ID (as
ExprIdand defaults to a new ExprId)
